Energy: Photovoltaics, Energy

Harvesting, Energy Storage, Fuel Cells

Improving Photovoltaic Conversion Using Spectral Conversion ............................................................................................ 67
Oxidative Chemical Vapor Deposition of Polymers for Printable and Flexible Photovoltaics............................................. 68
High-Efficiency Photovoltaic Devices using Nanocavity Array.............................................................................................. 69
Germanium-on-Silicon Heteroepitaxy for High-Efficiency Photovoltaic Devices.............................................................. 70
Solid-State Infrared-to-Visible Upconversion Sensitized by Colloidal Nanocrystals........................................................... 71
Solar Water Splitting with Metallic-Semiconductor Photonic Crystals................................................................................. 72
Optimizing Stability and Capacity for Lithium-Air Batteries................................................................................................... 73
Mechanical Stresses in Lithium Phosphorus Oxynitride-Coated Germanium Thin-Film Electrodes ............................... 74
Fabrication of Ruthenium Oxide-Coated Si Nanowire-Based Supercapacitors ................................................................. 75
Electro-Chemo-Mechanical Studies of Perovskite-Structured Mixed Ionic-Electronic
Conducting SrSn1-xFexO3-x/2+δ.................................................................................................................................................... 76
Utilization of BaSnO3 and Related Materials Systems for Transparent Conducting Electrodes........................................ 77
Oxygen Reduction Kinetics and Defect Chemistry of Layered Praseodymium Cuprate-Based Materials...................... 78
Investigation of Fuel Cell Cathode Performance in Solid Oxide Fuel Cells: Application
of Model Thin Film Structures..................................................................................................................................................... 79
Coupling of Oxide Catalyst Chemistry with Durability and Conversion Efficiency for Biomass Reforming..................... 80
Synthesis of CRUD and its Effects on Pool and Subcooled Flow Boiling.............................................................................. 81

MTL ANNUAL RESEARCH REPORT 2016 Energy 65

66 Energy MTL ANNUAL RESEARCH REPORT 2016
Improving Photovoltaic Conversion Using Spectral Conversion
D. M. Bierman, A. Lenert, W. R. Chan, B. Bhatia, I. Celanovic, M. Soljačić, E. N. Wang
Sponsorship: S3TEC, DOE

A single-junction PV cell generating electrical power complete thermalization of incident sunlight, a one-
through incident sunlight is limited by the Shock- dimensional Si/SiO2 photonic crystal for spectrally
ley-Queisser limit, which for any bandgap is far below selective thermal emission, and a tandem rugate/
the limiting efficiency of converting work from sun- interference filter. The PV cell used in the study is a
light. By introducing an intermediate component that low-bandgap InGaAsSb (Eg = 0.55 eV) cell.
absorbs sunlight and re-emits a tailored thermal radi- While absolute efficiencies are still low relative
ation spectrum towards a single junction PV cell, solar to more mature technologies, our peak measured
thermophotovoltaics (STPVs) promise to bridge the efficiency of 6.8% is above all others reported for
gap between these two disparate limits. STPVs. We have shown improved device conversion
For STPVs to succeed, a detailed understanding efficiency as well as reduced parasitic heat generation
of the energy flow via radiation and heat conduction as a result of the spectral converter. This is the first
is key. We have been working to enhance the time that the performance at the device level of a
performance of a single-junction PV by incorporating single junction PV cell has been enhanced due to the
a spectral converter. The spectral converter consists presence of a spectral converter.
of a multi-walled carbon nanotube (MW-CNT) area for

a b
2 cm
Spectral Converter

PV Cell

▲▲Figure 1: (a) Optical image of STPV device. Top surface has small MW-CNT area for photon absorption and tungsten-coated surface
to reduce heat loss through thermal emission. Bottom surface (not shown) produces spectrally selective thermal emission spectrum
through one-dimensional photonic crystal and rugate/interference filter. (b) Performance of device shows that STPVs can enhance
performance of PV cell by introducing intermediate absorption/re-emission process.

FURTHER READING

• W. Shockley and H. J. Queisser, “Detailed Balance Limit of Efficiency of P-N Junction Solar Cells,” Journal of Appl. Phys., vol. 32 no. 3, pp. 510, 1961.
• P. Würfel and U. Würfel, Physics of Solar Cells: From Basic Principles to Advanced Concepts, Hoboken: John Wiley & Sons, 2009.

MTL ANNUAL RESEARCH REPORT 2016 Energy 67

Oxidative Chemical Vapor Deposition of Polymers for Printable and Flexible
Photovoltaics
W. J. Jo, K. K. Gleason
Sponsorship: Eni S.p.A, Samsung Electronics, Institute for Soldier Nanotechnologies

A variety of solar energy conversion systems has from oCVD, various insoluble polymers are successfully
emerged as attractive candidates to establish fossil applied to OPVs. First, polyselenophene donor layers
fuel-free energy networks. Among these, organic so- are integrated into OPVs for the first time with 0.4 %
lar cells offer the promise of a lightweight, flexible, of power conversion efficiency. Second, ternary OPVs
large-area, and cost-effective photovoltaic technology. employing polythiophene donor layers has been
Traditionally, organic photovoltaic devices (OPVs) have realized to increase the power conversion efficiency
been fabricated with thermal evaporation or solu- up to 2.4%. Third, a new concept of neutral hole
tion-based techniques, but these two methods are suit- transporting layers (HTLs) is achieved by integrating
able for only low molecular weight or soluble materials. patterned Cl−doped poly(3,4-dimethoxy-thiophene)
In order to apply high molecular weight and insoluble (PDMT) HTLs into OPVs via oCVD. Due to this
polymers to OPVs, we explore the use of oxidative va- novel polymer’s neutrality, high transparency, good
por deposition (oCVD) for the polymer deposition. conductivity, and appropriate energy levels, the
As Figure 1 shows, oCVD is a solvent-free conformal power conversion efficiency and lifetime of OPVs
vacuum-based technique to enable thin-film fabrication are remarkably boosted compared to those of OPVs
of insoluble polymers at moderate vacuum (~ 0.1 Torr) depending on the commercial hole transporting
and low temperature (25–150°C). Moreover, oCVD carries polymer, acidic PEDOT:PSS [poly(3,4-ethylenedioxy-
the well-cited processing benefits of vacuum processing, thiophene):polystyrene sulfonate] (Figure 2). Finally,
such as parallel and sequential deposition, well-defined we are currently studying hyper-conductive PEDOT
thickness control, large-area uniformity, and inline for printable and flexible electronics by using a
integration with other standard vacuum processes vacuum-based polymer vapor printing technique –
(e.g., vacuum thermal evaporation). oxidative chemical vapor deposition (oCVD) combined
Based on the aforementioned technical advantages with in situ shadow masking.

▲▲Figure 1: Schematic of oCVD reactor. ▲▲Figure 2: Evolution of power conversion efficiency

with respect to storage time under N2 atmosphere.

FURTHER READING

• W. J. Jo, J. T. Nelson, S. Chang, V. Bulović, S. Gradečak, M. S. Strano, and K. K. Gleason, “Oxidative Chemical Vapor Deposition of Neutral Hole
Transporting Polymer for Enhanced Solar Cell Efficiency and Lifetime,” Advanced Materials, 2016.
• W. J. Jo, D. C. Borrelli, V. Bulović, and K. K. Gleason, “Photovoltaic Effect by Vapor-Printed Polyselenophene,” Organic Electronics, vol. 26, pp.
55-60, 2015.

68 Energy MTL ANNUAL RESEARCH REPORT 2016

High-Efficiency Photovoltaic Devices using Nanocavity Array
W. L. Kwan, A. El-Faer, X. Li, S.-G. Kim
Sponsorship: SUTD-MIT Program

We have demonstrated in our earlier results that met- The nanocavity array we fabricated can alleviate
al-dielectric nanocavity arrays can improve broad-band the need for thicker absorber film by enhancing
light absorption due to the enhancement in the cavity, broad-band absorption in the absorber material.
waveguide, and surface plasma polariton resonance Figure 1(a) shows the results of a finite-difference
modes around the nanocavity. This project aims to uti- time-domain (FDTD) simulation that the absorption
lize this absorber structure for applications in photo- in a poly(3-hexylthiophene-2,5-diyl):phenyl-C61-butyric
voltaic devices with soft absorber materials like organ- acid methyl ester (P3HT:PCBM) film is enhanced by
ic and perovskite materials, which have limited carrier the nanocavity structure (SEM image in Figure 1(b)).
lifetime and thereby require ultra-thin film thickness The enhanced absorption can also be clearly seen in
for efficiency charge extraction. However, light absorp- the photograph of a thin gold film deposited on the
tion is insufficient if the absorber layer is too thin, and nanocavity array in Figure 1(c). Most of the visible light
a compromise between complete light absorption and is absorbed, leaving a black appearance on the surface of
charge extraction is often needed, resulting in medio- the substrate. A prototype device is under fabrication to
cre energy harvesting efficiencies. demonstrate higher solar energy harvesting efficiency.

▲▲Figure 1: (a) FDTD simulation results on the absorption spectrum of P3HT:PCBM films on a flat surface and nanocavity
structure. (b) SEM image of Al2O3 nanocavity array coated with gold. The scale bar is 1mm. (c) Photograph of a silicon sub-
strate with Al2O3 nanocavity array coated with gold.

FURTHER READING

• J. B. Chou, D. P. Fenning, Y. Wang, M. A. M. Polanco, J. Hwang, A. EI-Faer, F. Sammoura, J. Viegas, M. Rasras, A. M. Kolpakl, Y. Shao-Hom, and
S. -G. Kim, “Broadband Photoelectric Hot Carrier Collection with Wafer-scale Metallic-semiconductor Photonic Crystals,” 42nd IEEE Photovoltaic
Specialist Conference (PVSC), New Orleans, LA, 2015.

MTL ANNUAL RESEARCH REPORT 2016 Energy 69

Germanium-on-Silicon Heteroepitaxy for High-Efficiency Photovoltaic Devices
B. R. Albert, L. C. Kimerling, J. Michel
Sponsorship: ARPA-E

While III-V based photovoltaic cells demonstrate from mesa corners. A staggered arrangement decreased
high energy-conversion efficiencies, their widespread the growth time until complete coalescence by at least
adoption is limited by the prohibitive cost per device 50% as compared to a regular gridded array. The rate
area. Substantial cost reduction of germanium lat- of overgrowth over isolated SiO2 lines was observed to
tice-matched III-V cells can be realized by replacement increase for smaller line widths up to 1.5 µm.
of the Ge substrate with a thin layer of Ge deposited Etch pit studies of coalesced structures employing
on silicon. To maintain efficient energy conversion, the faceted Ge growth around SiO2 walls arranged as grids
Ge-on-Si “virtual substrate” must be single crystalline and isolated lines indicated a local increase in the
with a threading dislocation density (TDD) low enough TDD in the vicinity of Ge film edges while decreasing
to not affect carrier lifetimes in the epitaxially deposit- to 1 x 107 cm-2 further away in the film. A significant
ed III-V layers. improvement in TDD reduction by avoidance of
Lateral overgrowth and film coalescence create a dislocation pile-up is expected by these same models
more suitable, planar Ge film for III-V growth in large if blanket Ge is instead grown, followed by etching
areas. Fabrication of a continuous Ge surface from and filling of trenches with poly-Ge separated by a
a patterned Ge film demands complete coalescence thin layer of SiO2. A 1-µm thick Ge film is expected to
during lateral overgrowth. Due to the high surface exhibit a TDD of 1 x 105 cm-2. At this density level, the
energy of the Ge/SiO2 interface, lateral overgrowth performance of high-efficiency III-V photovoltaic cells
does not readily occur. Ge mesa arrays were staggered will be unaffected.
to eliminate regions entirely dependent on overgrowth

▲▲Figure 1: Ge lateral overgrowth over SiO2 isolated walls. ▲▲Figure 2: Overgrowth at zero and negative concavity Ge
film perimeters.

70 Energy MTL ANNUAL RESEARCH REPORT 2016

Solid-State Infrared-to-Visible Upconversion Sensitized by Colloidal Nanocrystals
M. Wu, D. N. Congreve, M. W. B. Wilson, J. Jean, N. Geva, M. Welborn, T. Van Voorhis, V. Bulović, M. G. Bawendi, M. A. Baldo
Sponsorship: EFRC Center for Excitonics

Optical upconversion is a process that converts two been mostly conversion among visible wavelengths.
or more low-energy photons into a single high-energy Furthermore, the majority of them are in solution,
photon. Upconversion from the infrared to the visible while a solid-state architecture is necessary for solar
is useful in photovoltaics, photodetection, bioimaging and detection applications.
and three-dimensional displays. In photovoltaic appli- Here, we report a solid-state thin film for infrared-
cations specifically, an optical upconversion layer can to-visible upconversion that employs lead sulfide
capture sub-bandgap photons, increasing the efficien- colloidal nanocrystals as a sensitizer; see Figure 1.
cy of a conventional single-junction solar cell beyond Upconversion is achieved from pump wavelengths
the Shockley–Queisser limit. beyond λ = 1 μm to emission at λ = 610 nm (Figure 2). When
To upconvert light at relatively low intensities, excited at λ = 808 nm, two excitons in the sensitizer are
a promising approach is the sensitized triplet-triplet converted into one higher energy state in the emitter
annihilation (TTA). It utilizes a sensitizer and an at a yield of 1.2±0.2%. Upconversion efficiency reaches
annihilator. The sensitizer absorbs incident light the maximum at an absorbed intensity equivalent
and transfers the energy as spin-triplet excitons to less than one sun. We demonstrate that colloidal
to the annihilator. When two triplets meet in the nanocrystals are an attractive alternative to existing
annihilator, they form a single higher-energy spin- molecular sensitizers, given their small exchange
singlet exciton via TTA. Blue-shifted light is emitted splitting, wide wavelength tunability and broadband
when the singlets relax. It has been, however, difficult infrared absorption. This solid-state architecture for
to identify effective molecular sensitizers that absorb upconversion may prove useful for enhancing the
in the infrared. Efficient demonstrations to date have capabilities of solar cells and photodetectors.

Rubrene:0.5% DBP
80nm 1 1

Absorption, Excitation (arb.)

Photoluminescence (arb.)

Nanocrystal IR
sub monolayer 0.8 0.8
Glass
0.6 0.6
10mm
S ens itizer A nnihilator E mitter
P bS NC R ubrene DB P
0.4 0.4
T TA S1
S1 850nm
960 nm
E1 0.2 1010 nm 0.2
T 1=1.14eV
G 0 0
E xc itation E mis s ion 500 600 700 800 900 1000 1100 1200
Wavelength (nm)

▲▲Figure 1: Thin-film device structure (top) and energy diagram ▲▲Figure 2: Nanocrystal absorption labeled by first excitonic peaks,
(bottom) illustrating TTA-based upconversion sensitized by PbS DBP photoluminescence (red), and excitation spectra (purple
colloidal nanocrystals (NC). DBP: dibenzotetraphenylperiflan- crosses) for λ=1010nm NC. Inset: photograph of a device under
thene. λ=808nm excitation.

FURTHER READING

• M. Wu, D. N. Congreve, M. W. B. Wilson, T. Van Voorhis, V. Bulović, M. G. Bawendi, M. A. Baldo et al, “Solid-State Infrared-to-Visible Upconversion
Sensitized by Colloidal Nanocrystals,” Nature Photonics, vol. 10, pp. 31-34, 2016.
• T. N. Singh-Rachford and F. N. Castellano, “Photon Upconversion Based on Sensitized Triplet-Triplet Annihilation,” Coordination Chemistry
Reviews, vol. 254, pp. 2560-2573, 2010.
• T. F. Schulze and T. W. Schmidt, “Photochemical Upconversion: Present Status and Prospects for Its Application to Solar Energy Conversion,”
Energy & Environmental Science, vol. 8, pp. 103-125, 2015.

MTL ANNUAL RESEARCH REPORT 2016 Energy 71

Solar Water Splitting with Metallic-Semiconductor Photonic Crystals
X. H. Li, A. Elfaer, J. Chou, S.-G. Kim
Sponsorship: Masdar Flagship Program

The major challenges for current solar energy harvest- strated in Figure 1a, the generated hot electrons with
ing techniques are the limitations in high-efficiency en- enough energy to overcome the Au/TiO2 Schottky bar-
ergy conversion and large-scale energy storage. Direct- rier could be collected for water splitting.
ly converting solar energy into storable chemical fuels Figure 1b shows the micro-cavity arrays with depth
can solve these problems. A promising method is uti- of 1 μm and diameter of 500 nm. We deposited thin
lizing photoelectric hot electron generation to split wa- layers of TiO2 and Au to form the Schottky junction.
ter and produce hydrogen fuel. Titanium-oxide-based Finite-difference time-domain simulation results in
photocatalytic systems have been widely used in pho- Figure 2a show that thinner Au layers can promote
ton-driven hot electron generation. However, the effi- broadband light absorption, which might be beneficial
ciency of the present design is limited due to the low for hot electron generation. Experimental results of
absorption of visible light. Here we report hot electron the photoresponse in Figure 2b show a peak at around
collection by wafer-scale Au/TiO2 metallic-semicon- 590 nm, which is below the TiO2 bandgap. We also tried
ductor photonic crystals (MSPhC), with a broadband to achieve multi-band photoresponse by depositing
photoresponse below the bandgap of TiO2. Multiple gold nanorods on MSPhC. In order to understand the
absorption modes supported by the 2D micro-cavity generation and injection of hot electrons through
structure of the MSPhC extend the photon-metal in- plasmon decay, we are currently working on modeling
teraction time and fulfill broadband light absorption. and measuring momentum distribution of plasmon-
Surface plasmon absorption mode gives access to en- induced hot electrons in metal nanostructures such as
hanced electric field oscillation and hot electron gener- gold nanorods.
ation at the interface between Au and TiO2. As demon-

▲▲Figure 1: (a) Schematic of 2D micro-cavity arrays of ▲▲Figure 2: (a) Simulated absorption spectra of MSPhC
MSPhC. (b) Top view photo of MSPhC via SEM. Scale for various Au thickness. (b) Photoresponse of MSPhC
bars are 1 μm. centered at 590 nm.

FURTHER READING

• J. B. Chou, D. P. Fenning, Y. Wang, M. A. M. Polanco, J. Hwang, F. Sammoura, J. Viegas, M. Rasras, A. Kolpak, S. H. Yang, and S.-G. Kim, “Broadband
Photoelectric Hot Carrier Collection with Wafer-Scale Metallic-Semiconductor Photonic Crystals,” 42nd IEEE PVSC, New Orleans, LA, 2015.
• Y. Wang, J. B. Chou, and S.-G. Kim, “Simulation Study of Metallic Photonic Crystal for Enhanced Hot Electron Transfer in Electrochemical Cells,”
presented at MRS Fall Meeting, Boston, MA, 2015.
• A. Elfaer, Y. Wang, X. H. Li, J. B. Chou, and S.-G. Kim, “Gold Nanorods Coated Metallic Photonic Crystal for Enhanced Hot Electron Transfer in
Electrochemical Cells,” MRS Advances, FirstView, pp. 1-7, 2015.

72 Energy MTL ANNUAL RESEARCH REPORT 2016

Optimizing Stability and Capacity for Lithium-Air Batteries
T. Batcho, D. Kwabi, D. Perego, R. Omampuliyur, Y. Shao-Horn, C. V. Thompson
Sponsorship: Skoltech Center for Electrochemical Energy Storage

Lithium-air batteries hold promise for the next gen- passivate the carbon surface. These coated CNTs are
eration of electric vehicles and other applications. By capable of supporting Li2O2 growth (Figure 1). We
reacting oxygen directly with lithium ions to form are currently working on optimizing conductivity
Li2O2 on discharge, they can achieve energy densities of the deposited films and testing electrochemical
3-5 times higher than current lithium-ion batteries. performance during charge and cycling.
However, a number of challenges exist for implement- Another challenge in designing Li-O2 is obtaining
ing lithium-air batteries, including poor rate capability, optimal volumetric discharge capacity, which can be
poor cyclability, high overpotentials upon charging, achieved by promoting the growth of large toroidal
and electrode and electrolyte instability. We seek to ad- deposits of Li2O2 as opposed to thin films, which
dress these issues by developing new electrode materi- electrically passivate and cut off cell discharge
als and architectures and performing studies of Li2O2 prior to full void space filling of the electrode. We
formation under various discharge conditions. seek to study the mechanisms of nucleation and
Aligned arrays of carbon nanotubes (CNTs) growth in order to control these processes. For this
provide ideal conductive scaffolding materials for study we used carbon paper electrodes for greater
Li2O2, while having high void space and low mass. reproducibility and facility in modeling. We test these
CNTs of 5-10 nm in diameter are grown in aligned electrodes with potentiostatic discharges, which
forests on catalyst deposited silicon wafers. These use a fixed driving force. We can then adapt existing
forests can be delaminated and placed directly into our models for electrodeposition to our system to extract
cell. We observed near ideal gravimetric capacities and rates of surface nucleation and growth based on
high volumetric capacities. However, carbon has been current transients (Figure 2). By further studying the
found to decompose in lithium-air cells and promote dependence of these rates on solvent type, potential,
electrolyte decomposition. This leads to poor cycling and electrode surface, we can find optimal conditions
performance and high overpotentials on charge. To for greater cell capacity and better understand
avoid these effects, we coated materials such as TiN mechanisms of Li2O2 evolution.
onto CNTs using atomic layer deposition to chemically

▲▲Figure 1: SEM micrograph of toroidal Li2O2 formation on CNTs ▲▲Figure 2: Current transients from potentiostatic dis-
deposited with a coating of TiN. charges at a range of potentials in 0.1 M LiClO4 DMSO.
Peaks occur at longer times at higher potentials (lower
overpotentials), suggesting slower nucleation/growth rates.

FURTHER READING

• R. R. Mitchel, B. M. Gallant, Y. Shao-Horn, and C. V. Thompson, “Mechanisms of Morphological Evolution of Li2O2 Particles during Electrochemical
Growth,” Journal of Physical Chemistry Letters, vol. 4, no. 7, pp. 1060, March 2013.
• B. M. Gallant, R. R. Mitchell, D. G. Kwabi, J. Zhou, L. Zuin, C. V. Thompson, and Y. Shao-Horn, “Chemical and Morphological Changes of Li–O2
Battery Electrodes upon Cycling,” The Journal of Physical Chemistry C, vol. 116, no. 39, pp. 20800–20805, October 2012.

MTL ANNUAL RESEARCH REPORT 2016 Energy 73

Mechanical Stresses in Lithium Phosphorus Oxynitride-Coated Germanium Thin-
Film Electrodes
A. Al-Obeidi, D. Kramer, R. Moenig, C. V. Thompson
Sponsorship: SMART

In the electronics and health industry, there has been It was found that LiPON, a rigid solid electrolyte,
a strong trend toward miniaturized devices for use in suppresses morphological evolution and results
wearable electronics, medical implants, and wireless in reproducible cycle-to-cycle stress behavior. The
communication. The reduced energy consumption of repeatable behavior observed in LiPON-coated films
microsystems makes it possible to integrate microbat- allows more direct characterization of electrochemical
teries directly onto electronic chips. Solid-state micro- processes governing lithiation and delithiation. Cycling
batteries are ideally suited for such applications since at various rates (Figure 1) revealed that the lithiation
they can be integrated on microchips while offering capacity of coated electrodes increased at slower cycling
improved safety (no liquid electrolytes and thermal rates, saturating at about 1200 A h kg-1 when using rates
runaway), performance (higher voltages, wider range slower than 1 C. Cycling below 100 mV resulted in the
of operating temperatures), and lifetime. The simplest formation of c-Li15Ge4, which appeared as a sharp drop
and most common form of a solid-state battery is a in the compressive nominal stress to values close to
planar solid-state thin-film battery. For the anode, ger- zero (Figure 2, points 1-2). Overlithiation of this phase
manium is an ideal candidate since it offers large vol- resulted in a linear compressive increase in stress (Figure
umetric capacities (7366 A h l−1) compared to lithium 2, points 2-3). These results indicate that cLi15Ge4 has a
(2065 A h l−1) while being compatible with conventional higher density than its a-LixGe precursor. Delithiation
semiconductor processing techniques. However, use of c-Li15Ge4 seems to consist of two successive events:
of germanium is limited by the significant volumet- the formation of an intermediate phase followed by a
ric and structural changes that occur during cycling. rapid release of lithium from this intermediate phase,
In order to explore the relationship between electro- which resulted in the amorphization of the electrode.
chemistry and the mechanical stresses, in situ stress While crystalline Li15Ge4 develops a lower maximum
measurements on germanium thin-film electrodes nominal tensile stress than its amorphous counterpart,
coated with lithium phosphorus oxynitride (LiPON) extraction of lithium from cLi15Ge4 requires more energy
were performed. and therefore reduces the energy efficiency of a cell.

▲▲Figure 1: Nominal stress vs. capacity plots for a LiPON-coated 170-nm ger- ▲▲Figure 2: Nominal stress-capacity plots
manium film cycled galvanostatically (1 V → 5mV → 1 V) at different C rates. The for conditions that lead to the formation
data for the final scan at 0.2 C (purple) overlaps with that of the initial 0.2 C scan of crystalline Li15Ge4 of a LiPON-coated
(black). The 1 C rate corresponds to 115 µA cm-2. 170-nm-thick germanium electrode
cycled galvanostatically at 13.3 µA (i.e.,
0.2 C, 23 µA cm-2).

FURTHER READING

• Z. Choi, D. Kramer, and R. Mönig, “Correlation of Stress and Structural Evolution in Li4Ti5O12-based Electrodes for Lithium Ion Batteries,”
Journal of Power Sources, vol. 240, pp. 245, 2013.
• A. Al-Obeidi, D. Kramer, R. Mönig, and C. V. Thompson, “Mechanical Stresses and Crystallization of Lithium Phosphorous Oxynitride-coated
Germanium Electrodes During Lithiation and Delithiation,” Journal of Power Sources, vol. 306, pp. 817, 2016.

74 Energy MTL ANNUAL RESEARCH REPORT 2016

Fabrication of Ruthenium Oxide-Coated Si Nanowire-Based Supercapacitors
W. Zheng, Q. Chen, D. Wang, C. V. Thompson
Sponsorship: SMART

Supercapacitors are electrochemical devices that have the surface of ruthenium oxide, the high aspect ratio Si
high power density and long cycle life. Pseudo-capaci- nanowire structures coated with ruthenium oxide has
tors are a type of supercapacitor that involves revers- a high surface area of accessible ruthenium oxide per
ible surface reduction/oxidiation reactions. Among all area of substrate surface, which thus leads to a high
the pseudo-capacitive materials, ruthenium oxide is energy storage capacitance. We have developed an ALD
the most promising due to its high specific capacitance, process for coating of ruthenium oxide on Si nanowires
excellent cyclability, and high conductivity. While re- generated by MAAE. The composite structure showed a
searchers have been developing supercapacitors based con-tinuous coating of well-distributed particles (Figure
on ruthenium oxide or its composite with other materi- 1). High-resolution transmission electron microscopy
als such as carbon nanotubes (CNTs), there has been lit- characterization and x-ray diffraction analysis con-
tle study on ruthenium oxide-Si composite electrodes. firmed that most nanoparticles were in the form of
In earlier work, we demonstrated the feasibility of using elemental ruthenium. We are currently investigating
metal assisted chemical etching (MACE) to fabricate Si the electrochemical performance of this composite
nanowires for on-chip MOS capacitors. Here we use an material in an aqueous electrolyte using a three-
ordered vertical array of Si nanowires from a similar wet electrode setup. The preliminary data showed that
etching process named metal-assisted anodic etching the specific capacitance scaled well with the length of
(MAAE) to fabricate on-chip supercapacitors. silicon nanowires in this aqueous electrolyte (Figure 2).
Atomic layer deposition (ALD) is used to deposit Meanwhile, we are fabricating solid-state micro-
ruthenium oxide on silicon nanowires due to its conformal supercapacitors based on use of solid electrolytes. We
coating of high aspect ratio structures. The use of ALD are interested in studying the performance dependence
also provides precise control of the ruthenium oxide of the solid-state device on both Si nanowire aspect
film thickness. As pseudo-capacitive reactions occur at ratios and ALD cycle numbers.

▲▲Figure 2: Specific capacitance as function of Si nanowire

lengths. The black line shows linear fit to experimental data; the
blue line shows theoretical projection of the specific capacitance
▲▲Figure 1: TEM image of Si nanowires after ALD coating of with nanowire lengths.
ruthenium oxide.

FURTHER READING

• S. W. Chang, J. Oh, S. T. Boles, and C. V. Thompson, “Fabrication of Silicon Nanopillar-based Nanocapacitor Arrays,” Appl. Phys. Letters, vol. 96,
pp. 153108, 2010.
• S. W. Chang, V. P. Chuang, S. T. Boles, C. A. Ross, and C. V. Thompson, “Densely-packed Arrays of Ultrahigh-aspect-ratio Silicon Nanowire
Fabricated Using Block Copolymer Lithography and Metal-assisted Etching,” Adv. Funct. Mater., vol. 19, pp. 2495, 2009.
• C. Q. Lai, W. Zheng, W. K. Choi, and C. V. Thompson, “Metal Assisted Anodic Etching of Silicon,” Nanoscale, vol. 25, pp. 11123, 2015.

MTL ANNUAL RESEARCH REPORT 2016 Energy 75

Electro-Chemo-Mechanical Studies of Perovskite-Structured Mixed Ionic-Electronic
Conducting SrSn1-xFexO3-x/2+δ
C. S. Kim, S. R. Bishop, N. H. Perry, H. L. Tuller
Sponsorship: Skolkovo Foundation

High efficiency and fuel flexibility make solid oxide parallel to examine how changes in defect chemistry
fuel cells (SOFCs) attractive for conversion of fuels to and electronic band structure, associated with the
electricity. Reduced operating temperatures, desirable substitution of Ti by Sn, impact carrier density and
for reduced costs and extended operation, however, ultimately electrode performance. Bulk chemical
result in significant losses in efficiency. This loss has expansion was measured by dilatometry as a function
been traced primarily to slow cathode surface reaction of oxygen partial pressure, while surface kinetics
kinetics. In this work, we extend previous studies on were examined by means of AC impedance spectroscopy.
the promising mixed ionic and electronic conducting The electro-chemo-mechanical properties of SSF were
perovskite-structured SrTi1-xFexO3-x/2+δ (STF) materials found not to differ significantly from the corresponding
system whose exchange kinetics were correlated with composition in STF. Key thermodynamic and kinetic
the minority electron charge density by replacing Ti parameters for SrSn0.65Fe0.35O2.825+δ (SSF35) were derived,
with Sn, due to its distinct band structure and higher including the reduction enthalpy, electronic band gap,
electron mobility. anion Frenkel enthalpy, oxygen vacancy migration
Oxygen nonstoichiometry and the defect chemistry energy, and electron and hole mobilities. Though slightly
of the SrSn1-xFexO3-x/2+δ (SSF) system were examined shifted by the larger size of Sn, the defect equilibria and
by thermogravimetry as a function of oxygen partial the cathode area specific resistance differed only in a
pressure in the temperature range of 973-1273 K. limited way from that in STF. This was attributed to
Marginally higher reducibility was observed compared properties being largely dominated by Fe and not by the
to corresponding compositions in the STF system. substitution of Ti with Sn.
The bulk electrical conductivity was measured in

▲▲Figure 1: Oxygen nonstoichiometry δ as a function of ▲▲Figure 2: Comparison of temperature dependence of

pO2 and temperature for SSF35. area specific resistance (ASR) of SSF35 and STF35 thin
film electrodes.

FURTHER READING

• W. Jung and H. L. Tuller, “A New Model Describing Solid Oxide Fuel Cell Cathode Kinetics: Model Thin Film SrTi1-xFexO3-δ Mixed Conducting
Oxides, a Case Study,” Advanced Energy Materials, vol. 1, pp. 1184–1191, 2011.

76 Energy MTL ANNUAL RESEARCH REPORT 2016

Utilization of BaSnO3 and Related Materials Systems for Transparent Conducting
Electrodes
M. Campion, S. R. Bishop, H. L. Tuller
Sponsorship: NSF

Efficient transparent electrode materials are vital for and material properties of BaSnO3 in order to better
applications in smart window, LED display, and so- establish the consistent and controllable use of BaSnO3
lar cell technologies. These materials must possess a as a transparent electrode. To accomplish these goals,
wide band gap for minimal optical absorption in the methods such as in-situ resistance and impedance
visible spectrum while maintaining a high electrical monitoring during annealing will be applied. In
conductivity. Tin-doped indium oxide (ITO) has been addition, a variety of novel methods such as the in-
the industry standard for transparent electrodes, but situ monitoring of optical transmission (shown in
the use of the rare element indium has led to a search Figure 1) during annealing and the in-situ monitoring
for better material alternatives. BaSnO3 represents a of resistance during physical vapor deposition will be
promising alternative due to its high electron mobility utilized to investigate BaSnO3. Direct measurements
and resistance to property degradation under oxidizing of the key constants for the thermodynamics and
conditions, but the mechanisms by which processing kinetics of oxidation in donor-doped BaSnO3 will be
conditions and defect chemistry affect the final mate- experimentally determined for the first time. This
rial properties are not well understood. increase in understanding will provide a predictive
This work seeks to better understand the model for determining the optical properties, carrier
relationships between processing, defect chemistry, concentrations, and electron mobilities in BaSnO3.

▲▲Figure 1: Schematic of experimental setup to be used for simultaneous in-situ measurement of the optical transmission and
electrical conductivity of thin film BaSnO3 samples during annealing under controlled atmosphere and temperature.

FURTHER READING

• D. O. Scanlon, “Defect Engineering of BaSnO3 for High-performance Transparent Conducting Oxide Applications,” Phys. Rev. vol. 87, no. 16, pp.
161201, April 2013.
• J. J. Kim, S. R. Bishop, N. J. Thompson, D. Chen, and H. L. Tuller, “Investigation of Nonstoichiometry in Oxide Thin Films by Simultaneous in Situ
Optical Absorption and Chemical Capacitance Measurements: Pr-Doped Ceria, a Case Study,” Chemistry of Materials, vol. 26, pp. 1374-1379, 2014.

MTL ANNUAL RESEARCH REPORT 2016 Energy 77

Oxygen Reduction Kinetics and Defect Chemistry of Layered Praseodymium
Cuprate-Based Materials
C. S. Kim, K. Mukherjee, S. R. Bishop, H. L. Tuller in collaboration with Y. Hayamizu, S. Istomin
Sponsorship: Skolkovo Foundation

Layered cuprate compounds with mixed ionic electronic data to defect models enables the derivation of
conductivity are promising candidate materials for cath- thermodynamic parameters as well as defect carrier
odes in intermediate-temperature solid oxide fuel cells. concentrations. Dopants added to praseodymium
There have been reports of anisotropic oxygen diffusion cuprate were found to considerably extend the range of
in materials with the K2NiF4 (T) and Nd2CuO4 (T’) crys- oxygen nonstoichiometry. Thins films of corresponding
tal structures, with facile transport along the rock-salt compounds were prepared by pulsed-laser deposition
layers. These material systems also exhibit anisotropic onto single-crystal YSZ substrates. Alloys with the
thermal and chemical expansion properties, potential- same crystal structure and doping but with different
ly important for long-term device stability. However, in film orientations were also successfully synthesized
practice, it is difficult to independently control the crys- through the use of seed layers on YSZ (Figure 1). Using
tal structure, doping, and grain orientation to under- electrochemical impedance spectroscopy to measure
stand their effects on cathode performance. Additional- the area-specific resistance (ASR), we found a significant
ly, the lanthanide cuprates exhibit a third phase (T*), a improvement in the oxygen surface-exchange rate as
hybrid of the T and T’ phases, with important implica- a result of both donor and acceptor doping (Figure 2).
tions for the ionic and electronic conductivity. However, the activation energies are very different,
In this work, we aim to understand the correlation indicative of different rate-determining steps in each
between defect chemistry and surface exchange case. Also, contrary to expectations, the effect of
kinetics. Oxygen nonstoichiometry of Pr2CuO4 with film orientation (fast oxygen diffusion axis vs. slow
varying amounts of Sr and Ce doping is studied as a diffusion axis) had only a relatively weak effect on the
function of oxygen partial pressure and temperature ASR. The reason for this weak correlation between the
by thermogravimetry. Fitting the nonstoichiometry film orientation and the ASR is under investigation.

▲▲Figure 1: Cross-sectional TEM of a Pr2CuO4 thin film grown ▲▲Figure 2: Oxygen surface exchange as a function of in-
epitaxially on a YSZ substrate. The columnar grains are a result verse temperature in Pr2CuO4 thin films as determined
of tetragonal film on cubic substrate resulting in two possible from the area-specific resistance by electrochemical im-
in-plane orientations. pedance spectroscopy. The activation energies are listed.

FURTHER READING

• M. Burriel, G. Garcia, J. Santiso, J. A. Kilner, R. J. Chater, and S. J. Skinner, “Anisotropic Oxygen Diffusion Properties in Epitaxial Thin Films of
La2NiO4+δ,” Journal of Materials Chemistry, vol. 18, pp. 416-422, 2008.
• G. N. Mazo, Y. A. Mamaev, M. Z. Galin, M. S. Kaluzhskikh, and A. K. Ivanov-Schitz, “Structural and Transport Properties of the Layered Cuprate
Pr2CuO4,” Inorganic Materials, vol. 47, pp. 1218–1226, 2011.
• H. Y. Hwang, S.-W. Cheong, A. S. Cooper, L. W. Rupp Jr., B. Batlogg, and G. H. Kwei, “Crystallographic Evolution, T’→T*→T, in Pr2-xSrxCuO4-δ,” Physica
C: Superconductivity, vol. 192, pp. 362–371, 1992.

78 Energy MTL ANNUAL RESEARCH REPORT 2016

Investigation of Fuel Cell Cathode Performance in Solid Oxide Fuel Cells:
Application of Model Thin Film Structures
J. J. Kim, D. Chen, N. H. Perry, S. R. Bishop, H. L. Tuller
Sponsorship: DOE

An improved fundamental understanding of oxygen reducing/oxidizing conditions, the ability to diagnose a

nonstoichiometry (δ) and surface exchange kinetics in material’s behavior in situ is, therefore, important.
solid oxide fuel cell (SOFC) cathodes is considered criti- Our group recently demonstrated that δ in
cal for achieving enhanced device performance and lon- Pr0.1Ce0.9O2-δ (10 PCO) thin films could be reliably
gevity, especially at reduced operating temperatures. Al- derived by utilizing chemical capacitance extracted
though numerous research activities have been focused from electrochemical impedance spectroscopy (EIS)
on elucidating the oxygen reduction reaction (ORR) measurements. Furthermore, we introduced a non-
mechanisms at the cathode, their conclusions remain contact optical means for in situ recording of transient
unsatisfactory and controversial. The ORR at mixed redox kinetics, as well as the equilibrium Pr oxidation
conducting oxide thin film cathodes consists of oxygen state and, in turn, δ in 10 PCO thin films, by monitoring
adsorption, dissociation, charge-transfer, incorporation, the change in absorption spectra upon change in pO2 or
and migration of charge carriers. The kinetic parame- temperature. Recently, these optical techniques have
ters associated with the overall ORR, such as the diffu- been extended by our group to examination of the
sion coefficient (D) and surface exchange coefficient (k), perovskite SOFC electrode Sr(Ti,Fe)O3-δ. In this study, we
are strongly influenced by δ in the oxides. On the other are investigating cathode kinetics and nonstoichiometry
hand, oxygen defect generation is often associated with of two model oxide thin films, Sr(Ti,Fe)O3-δ (STF) with
valence changes in the transition metal or rare earth La doping and (Pr,Ce)O2-δ (PCO), by simultaneously
ions within the oxides and corresponding changes in utilizing in situ and in operando optical absorption
lattice constant (chemical expansion). This phenom- spectroscopy and EIS as a function of temperature,
enon may lead to stresses sufficient to support crack pO2 and electrical potential. We are also investigating
initiation and/or delamination, impacting the device’s changes in surface chemistry and their impact on
long-term stability. Because many advanced oxide ma- electrode impedance by atomic force microscopy (AFM),
terials used in SOFCs experience significant changes in x-ray photoelectron spectroscopy (XPS), and low-energy
δ during operation at elevated temperatures and under ion scattering spectroscopy (LEIS).

FURTHER READING

• D. Chen, S. R. Bishop, and H. L. Tuller, “Non-stoichiometry in Oxide Thin Films Operating Under Anodic Conditions: A Chemical Capacitance
Study of the Praseodymium-cerium Oxide System,” Chemistry of Materials, vol. 26, no. 22, pp. 6622-6627, 2014.
• N. H. Perry, D. Pergolesi, S. R. Bishop, and H. L. Tuller, “Defect Chemistry and Surface Oxygen Exchange Kinetics of La-doped Sr(Ti,Fe)O3-a in
Oxygen-rich Atmospheres,” Solid State Ionics, vol. 273, pp. 18-24, 2015.
• J. J. Kim, S. R. Bishop, N. J. Thompson, D. Chen, and H. L. Tuller, “Investigation of Nonstoichiometry in Oxide Thin Films by Simultaneous in Situ
Optical Absorption and Chemical Capacitance Measurements: Pr-doped ceria, a Case Study,” Chem. Mater., vol. 26, pp. 1374-1379, January 2014.

MTL ANNUAL RESEARCH REPORT 2016 Energy 79

Coupling of Oxide Catalyst Chemistry with Durability and Conversion Efficiency
for Biomass Reforming
V. Mortola, S. R. Bishop, H. L. Tuller in collaboration with M. Shetty, Y. Roman
Sponsorship: Basic Energy Sciences/DOE

The transportation sector currently relies heavily on

non-renewable fossil-based fuels to meet its energy
requirements. Alternative energy sources need to be
developed to meet the projected energy demand in the
future in a sustainable and environmentally conscious
manner. Bio-oils, generated via biomass fast pyrolysis,
represent an attractive avenue for the production of
renewable fuels and chemicals. We are investigating
MoO3-δ, a promising catalyst for bio-oil upgrading, that
selectively transforms various phenolic compounds
into aromatic hydrocarbons with high yields using low
H2 pressures by hydrodeoxygenation (HDO). While it is
clear that both ionic (oxygen vacancies) and electronic
defects (e.g., Mo5+/3+), in both the oxide catalyst and ox-
ide support (e.g., TiO2, ZrO2 and Al2O3), play controlling
roles in the catalytic reactions, surprisingly little is
known about how operating conditions (temperature,
dopants/impurities, and atmosphere (pO2, pH2O, pH2,
carbon activity, etc.)) impact the concentration or the
diffusivities of these key defects, or how they influence
the reaction kinetics. In this project, we combine catal-
ysis with defect chemical and impedance spectroscopic
approaches to achieve a deeper understanding of the
role that defects in metal oxide catalysts and supports
play in influencing catalytic selectivity, efficiency, sta-
bility, and reaction rates.

FURTHER READING

• T. Prasomsri, M. Shetty, K. Murugappan, and Y. Roman-Leshkov, “Insights into the Catalytic Activity and Surface Modification of MoO3 During
the Hydrodeoxygenation of Lignin-derived Model Compounds into Aromatic Hydrocarbons Under Low Hydrogen Pressures,” Energy Environ.
Sci., vol. 7, pp. 2660, 2014.
• T. Prasomsri, T. Nimmanwudipong, and Y. Roman-Leshkov, “Effective Hydrodeoxygenation of Biomass-derived Oxygenates into Unsaturated
Hydrocarbons by MoO3 Using Low H2 Pressures,” Energy Environ. Sci., vol. 6, pp. 1732, 2013.

80 Energy MTL ANNUAL RESEARCH REPORT 2016

Synthesis of CRUD and its Effects on Pool and Subcooled Flow Boiling
C. P. Coyle, B. A. Phillips, S. P. Lowder, J. Buongiorno, M. Bucci, T. McKrell
Sponsorship: DOE, Office of Nuclear Energy

Chalk River Unidentified Deposits (CRUD) are a nat- photoresist posts, and confirm film thickness using
urally occurring porous, hydrophilic layer that forms the Dektak profilometer. Features such as thickness,
on fuel rods during nuclear reactor operation. Unique wettability, pore size, and chimney diameter and pitch
features of these deposits are the characteristic boiling were verified. During pool and flow boiling testing,
chimneys, as shown in Figure 1. It has been hypothe- IR thermography and high-speed video were used to
sized that the presence of these chimneys, by provid- obtain temperature profiles of the active heater area
ing a clear path for vapor escape, can further enhance to quantify properties such as heat transfer coefficient,
boiling properties such as critical heat flux and heat nucleation site density, bubble departure frequency,
transfer coefficient. and bubble departure diameter.
An investigation of such effects has been conducted Data from pool and subcooled flow boiling tests
by preparing a porous, hydrophilic layer with boiling has shown that the heat transfer coefficient increases
chimneys on indium-tin-oxide-coated sapphire heaters. with increasing layer thickness and chimney diameter
A porous matrix emulating CRUD, shown in Figure 2, while the chimney pitch has relatively no effect. The
was created using layer-by-layer deposition of 100-nm observed increase results from greater nucleation site
silica nanoparticles to form porous, hydrophilic thick densities and greater bubble departure frequencies,
films. Photolithography was used to manufacture posts meaning the surface is able to remove more heat
that were then dissolved to create characteristic boiling through the creation of more bubbles per unit area and
chimneys. MTL facilities were used to deposit gold pads per unit time. The bubble parameters also followed
on the heater, plasma clean heaters, create and remove expected trends with mass flux and imposed heat flux.

▲▲Figure 1: SEM image of actual reactor CRUD ▲▲Figure 2: SEM image of synthetic CRUD with a zoom-in of the
(note the micro-scale boiling chimneys). nanoporous CRUD matrix in between the boiling chimneys.

MTL ANNUAL RESEARCH REPORT 2016 Energy 81

82 Energy MTL ANNUAL RESEARCH REPORT 2016