Iron pyrite (fool's gold, FeS₂) is a promising earth abundant and environmentally benign semiconductor material that shows promise as a strong and broad absorber for photovoltaics and high energy density cathode material for batteries. However, controlling FeS₂ nanocrystal formation (composition, size, shape, stoichiometry, etc.) and defect mitigation still remains a challenge. These problems represent significant limitations in the ability to control electrical, optical and electrochemical properties to exploit pyrite's full potential for sustainable energy applications. Here, we report a symmetry-defying oriented attachment FeS₂ nanocrystal growth by examining the nanostructure evolution and recrystallization to uncover how the shape, size and defects of FeS₂ nanocrystals changes during growth. It is demonstrated that a well-controlled reaction temperature and annealing time results in polycrystal-to-monocrystal formation and defect annihilation, which correlates with the performance of photoresponse devices. This knowledge opens up a new tactic to address pyrite's known defect problems.
Iron pyrite (fool's gold, FeS₂) is a promising earth abundant and environmentally benign semiconductor material that shows promise as a strong and broad absorber for photovoltaics and high energy density cathode material for batteries. However, controlling FeS₂ nanocrystal formation (composition, size, shape, stoichiometry, etc.) and defect mitigation still remains a challenge. These problems represent significant limitations in the ability to control electrical, optical and electrochemical properties to exploit pyrite's full potential for sustainable energy applications. Here, we report a symmetry-defying oriented attachment FeS₂ nanocrystal growth by examining the nanostructure evolution and recrystallization to uncover how the shape, size and defects of FeS₂ nanocrystals changes during growth. It is demonstrated that a well-controlled reaction temperature and annealing time results in polycrystal-to-monocrystal formation and defect annihilation, which correlates with the performance of photoresponse devices. This knowledge opens up a new tactic to address pyrite's known defect problems.
Iron pyrite (fool's gold, FeS2) is an eco-friendly material that is abundant
in nature and is extremely promising for use as an active layer in photovoltaics,
photoelectrochemical cells, broad spectral photodetectors and cathode material for
batteries1234. Pyrite boasts a strong light absorption (α >
105 cm−1), a suitable band gap of Eg =
0.95 eV5. and an adequate minority carrier diffusion length
(100–1000 nm)67, and more importantly, exhibits non-toxicity
and near-infinite elemental abundance. Enhancing its excellent properties requires basic
research on the controlled growth of pyrite, such as shape, size and stoichiometry. Controlled
preparation of FeS2 nanocrystals with specific sizes and shape has been
investigated in studies, involving the synthesis of zero dimensional (0D) nanoparticles8, one dimensional nanowires (1D)9, two dimensional (2D) thin
hexagonal sheets1 and three dimensional (3D) nanocubes10111213. Synthesis-by-design and understanding underlying growth mechanisms
is an especially important tool for targeted energy harvesting or storage applications.
Therefore, tailoring the size, shape and properties of pyrite nanostructures is a major
challenge that must be overcome before use in practical applications.In the past, as classical crystal growth kinetics models, LaMer and Ostwald ripening (OR)
theories have been widely used for the controlled synthesis of various colloidal
nanoparticles, in which the initial nucleation and growth can be explained by the
Gibbs-Thompson law1415. More recently, a novel growth process called
Oriented-Attachment (OA) has been identified which appears to be a unique mechanism during the
development of nanoscale materials161718192021. The aggregation
controlled OA provides an important route by which nanocrystals grow, an explanation of how
defects (dislocation) are formed and unique crystal morphologies, often symmetry-defying, can
be produced. The OA process was first described by Penn and Banfield et. al2223. Recently, Tang and Kotov reported the controllable synthesis of inorganic nanocrystal
materials using the self-assembly based OA mechanism2425. The interaction
force among particles plays an important role in the OA process, such as dipole-dipole
interaction, electrostatic repulsion, van der Waals interaction, and hydrophobic
attraction252627. The basics of the OA process are (1) primary
nano-clusters or particles aggregate, (2) a rotational step to achieve collision of higher
energy surfaces occurs, (3) removal of surfactants or absorbates, and finally (4) coherence is
achieved by combination of the high surface energy facets into a single crystal that results
in the reduction the overall surface energy of the particle. This coherence, while
thermodynamically favorable, may also create line and plane defects and twining. These defects
can lead to different properties of the material and give clues to crafting optimized
FeS2 nanocrystals for device applications and can be used to explain poor
performance of previous attempts at pyrite solar cell devices2.In this study, a novel growth mechanism of FeS2 pyrite nanocrystals is presented.
The new process exhibits a combination of LaMer theory for the initial quantum dot seeds
followed by OA growth to create the shape, size and crystallinity of the FeS2
nanocrystals. The OA growth is observed in creation of four different shapes of
FeS2 nanocrystals (cube, sheet, hexagonal plate and sphere) implying this is a
dominant mechanism for FeS2 nanostructures. Observing an OA growth mechanism could
offer insight into pyrite's known problems that have been attributed to vacancies and
crystal defects that hold it back as a highly promising photovoltaic material28. High-resolution transmission electron microscopy (HRTEM) images show the progression from
initial seeds to final monocrystal phase. To our knowledge, the OA growth has not been
reported utilizing a hot-injection method, as usually a precipitation method is used to create
the initial seeds and the final crystals. Finally, it is shown that FeS2 sheets
created from the OA growth process can be integrated into a photodetector device and can be
used as a probe for defect mitigation, and more importantly, shows the extent of
recrystalization's effect on optoelectronic performance.In the following report, evidence will first be presented for the OA growth in multi-shaped
FeS2 nanocrystals and a proposed reasoning for final shape created in the
nanocrystals. Characterization of FeS2 nanocrystals will be presented next followed
by tuning of the nanocrystal size by utilizing OA kinetics. Finally, the performance of
photodetector devices created out of FeS2 pyrite nanosheets will be presented and
analyzed.
Results
The initial step in synthesis of FeS2 nanocrystals consists of the creation of
FeS2 quantum dot (QD) seeds. QD formations are realized by a rapid
hot-injection of sulfur into an iron precursor solution, quickly creating QDs which show an
average diameter of 2 nm with a narrow size distribution (Figure
1 and Supplementary Fig. S1) and create a transparent deep blue
solution when dissolved in chloroform. Figure 1b shows the optical
absorption spectrum of FeS2 QDs which exhibit strong quantum confinement and
well-defined excitonic features, that match well with a previous report29.
The OA process then proceeds utilizing the QDs as primary particle seeds. By controlling the
injection temperature, different surface facet-rich nanocrystals can be obtained, which
directs the collision or the attachment direction and thus control the cube or
symmetry-defying sheet growth.
Figure 1
TEM images of FeS2 QDs (a), inset is the high resolution TEM image, and
UV-Vis absorbance (b).
Since different surface facets of FeS2 QD seeds exhibit different surface
energy, anisotropic OA growth is realized by the combination of energetically unfavorable
surface facets which will reduce the overall energy of the formed FeS2
nanocrystals. After the aggregation occurs (See Supplementary Fig. S2),
the OA process continues with the formation of a polycrystalline structure followed by a
recrystalization to a monocrystal. TEM images of each step for FeS2 nanocube
formation are presented in Fig. 2a–2d. Optical absorption
spectrum tracking structure changes are presented in Supplementary Fig.
S3. Note that by the aggregation step, a cube-like shape can already be seen being
formed (Fig. 2b). The OA growth is defined by the material's
symmetry and the surface facets of FeS2 which exhibit the lowest energy30. By increasing the injection temperature from 393 K to 418 K,
thin FeS2 pyrite {100} nanosheets are formed for the first time by the OA
mechanism (Fig. 2e–2h). The small seeds can be seen within the
sheet-like matrix (Fig. 2f), reminiscent of PbS sheets formed by OA
growth31. In the case of the nanosheets, it is seen that final sheets grow
thicker from aging (Fig. 2h). Supplementary Figure
S4 presents the thickening evidence through the TEM cross section of the sheets at
different growth time.
Figure 2
Sequences of TEM images show the detail of the attachment process.
(a) FeS2 QD seeds; (b) seed collision; (c) seed coalescence; (d)
recrystallization process from polycrystal to monocrystal. (e–h) FeS2
seeds evolved into single crystal nanosheet by coalescence and recrystallization process
(inset of Fig. 2h shows Fast Fourier Transform of nanosheet).
We interpret the symmetry-defying OA growth mechanism of our FeS2 nanocrystals
based on the thermodynamic stability of different surface facets predicted by Barnard and
Russo30. In their work, it is shown a truncated FeS2 nanocluster
of 5 nm is made up with 6 {100}, 8 {111} and 12 {110} surface facets. Figure 3 presents a depiction of these nanoclusters and the paths to different
shape formation seen in this study. In the case of cube growth (path A), a relatively larger
FeS2 QD seed is created at a lower injection temperature, which results in
mainly {100} surface planes being formed. The FeS2 QD seeds are stabilized by the
OA preferentially along {100} facets to form cubic FeS2 nanocrystals with {100}
surface planes. Regarding the FeS2 nanosheet formation, creation of relatively
smaller crystallites with higher {110} surface area explains the in-plane attachment.
FeS2 QD seeds with {110}-rich surfaces are created when the temperature of the
injection is increased. We interpret the thin FeS2 nanosheet formation by the
aggregation of the seeds through the {110} surface plane, shown in Fig.
3, pathway B. Since {110} surface facets of FeS2 seeds have higher
surface energy, they are preferentially consumed by the in-plane 2D attachment, resulting in
the FeS2 nanosheet formation. Conversion to the thicker sheet structures most
likely occurs through attachment of the sheets prevalent {100} surface, as the planar
dimension does not change in size, shown in Supplementary Fig. S4. The
ability to control initial seeds and their surface facets will prove valuable in extracting
the obtainable shapes of FeS2 nanocrystals (see Supplementary Fig.
S5), which have shown vastly different properties in optoelectronic and
electrochemical devices by us (this will be discussed in a future report).
Figure 3
Schematic illustration of the cubic (pathway A) and sheet (pathway B) formation of
FeS2 nanocrystals.
It has been observed and explained kinetically that in the OA growth model, higher growth
temperature leads to smaller particles due to the extra energy allowing for easier
de-adsorption of the particles during the collision step of the OA based growth32. The OA controlled tunability of FeS2 nano-crystal dimensions is
confirmed by varying growth temperature of the cubic synthetic route. Figure
4a–4c shows TEM images of FeS2 nanocrystals when the growth
temperature was 493, 523 and 543 K, respectively. Quantitatively, it is seen that as
the growth temperature increases, the average size of the final FeS2 nanocrystal
decreases from 64 nm, 43 nm, and 23 nm, respectively, providing
additional evidence of an OA controlled growth mechanism. Another key difference between OR
and OA growth in nanocrystals is particle size dependence on the growth time. In OR, as
stated above, bigger particles grow at the expense of smaller particles, making size
increase as time progresses. In OA, the particles attach to create a more stable particle,
and then usually cease to grow afterwards (there exists cases where after an OA step, OR
takes over and some growth still occurs). This leads to a stagnation of size after the OA
growth has taken place. In this study, there exists a point where the FeS2
nanostructures stop growing in size. Supplementary Figure S6 shows the
size of cubic structures at 40 min and 120 min into the synthesis and there is
no observed change in the overall sizes. Controlling the size of FeS2
nanoparticles is an important goal, as it has been stated that only a 40 nm film is
required in devices due to the material outstanding absorption coefficient12.
Figure 4
FeS2 cubic nanocrystals at different growth temperature.
(a) FeS2 nanocube at 493 K growth, (b) FeS2 nanocube at
523 K growth and (c) FeS2 nanocube at 546 K growth. (d) The
kinetic energy (KE) and the dipole-dipole potentials as the function of
reaction temperature for FeS2 nanocrystal growth.
The existence of different size and shape of FeS2 nanocrystals suggests
different collision and coalescence behavior of FeS2 seed crystallites. In the OA
growth process, the reaction temperature dominates the collision and the coalescence which
is attributed to the particle's medium- and short-range interactions, such as Van der
Waals forces and dipole-dipole interaction forces. Van der Waals forces are estimated to be
less than 0.5 RT, which is not enough to stabilize superstructures under ambient
conditions33. The force capable of producing FeS2 polycrystals
is thus believed to be the long range dipole-dipole attraction. The energy of dipole
attraction between FeS2 QD seeds can be calculated using the classical formula
=
−μ2/2πεor(r2
− dNP2). Estimating the center-to-center interdipolar
separation r to be 2.6 nm, the FeS2 QD seed diameter to be
dNP = 2 nm and taking the dipole moment for this size μ =
17.6D, the energy of dipole attraction is equal to 5.2 kJ/mole34. In the weakly flocculated colloidal state, the dipole-dipole potential can
also be expressed as a function of temperature T. When dipole-dipole potentials
and kinetic energy (KE) are plotted as a function of T (Fig. 4 (d)),
an intersection represents a critical temperature, Tc of the system at
416 K. Tc represents when the thermal energy exceeds the attractive
potential energy among FeS2 seeds. If the reaction temperature is lower than the
Tc, the attractive dipole-dipole potential energy dominates the OA process by
coalescence. Once the reaction temperature exceeds Tc, the KE will control the OA
growth, which is dictated by the collision. The size of FeS2 nanocrystals will be
controlled by reaction time in the coalescence state. In our FeS2 synthesis, we
control the coalescence and collision to yield FeS2 nanocubes with different
sizes, by tuning the heating rate and thus the reaction time within the coalescence state
(Fig. 4a–4c).While the shape and size control is an important goal in the FeS2 system, defect
mitigation may be the most crucial aspect in achieving optimal FeS2
nanostructures. It has been widely accepted that the defects (such as, surface states,
dislocations, twins, etc) of FeS2 nanocrystals dictate their optoelectronic and
electrochemical applications, therefore a strategy to achieve high quality crystalline
FeS2 needs to be identified. Polycrystalline-to-monocrystalline conversion of
the OA growth can be utilized to create highly crystalline FeS2 nanocrystals.
Figure 5a shows a HRTEM image of a FeS2 nanocube at
40 min into the synthesis. It can be seen that different domains (outlined by lines)
exist, while stacking faults can clearly be seen (highlighted by arrows) due to the
collision of the OA growth process. These defects are detrimental to material quality as
they act as the charge recombination centers for excitons and need to be eliminated to
create optimal solar cell devices. Upon greater lengths of aging time in the same pyrite
solution, it is seen that these defects are eventually eliminated. Figure
5b shows a monocrystalline cubic FeS2 nanocrystal aged for
120 min and the inset shows [100] growth diffraction pattern. This suggests
that longer aging times will be beneficial for FeS2 nanomaterial, due to the
stagnation of the OA controlled FeS2 growth, the longer aging times should not
interfere with shape/size.
Figure 5
HRTEM images of one cubic FeS2 nanocrystal recrystallization
process.
(a) A polycrystal FeS2 nanocube at 40 min into synthesis. Different
domains are separated by stacking faults (outlined as arrows) due to the collision of
the OA growth. (b) A monocrystal FeS2 nanocube after aging 120 min and
the inset shows the {100} diffraction pattern.
Discussion
As stated above, it is widely agreed that defects in pyrite material is the limiting factor
for performance of devices2353637. In this study, it is seen that the
pyrite particles eventually reach a maximum size, and then begin to convert from
poly-crystalline to mono-crystalline, which will reduce the defects that are caused by the
OA mechanism. To test this hypothesis, a series of photodetector devices were fabricated,
using FeS2 nanocrystals with varying aging times to examine the effect of
crystallinity on the device performance. The time-dependent photoresponse of FeS2
nanosheets are shown in Fig. 6 (a). The current difference between
irradiation (light on) and dark (light off) is clearly enhanced by increasing the aging time
of the FeS2 nanosheets. The figure of merit we use to compare photodetector
performance is the normalized detectivity (D*)38. D* values of the
FeS2 nanosheet devices are 5.84 × 1010, 8.60 ×
1010, and 1.85 × 1011 Jones, corresponding to
10 min, 40 min, and 240 min aging time of nanosheets, respectively
(shown in Fig. 6(b)). Since FeS2 nanosheets demonstrate a
strong absorbance in the near infrared (NIR) wavelength, they could work as the NIR
photodetector. Figure 6b shows the performance under 1000 nm
illumination, which confirms excellent NIR performance and again demonstrates the effect of
crystallinity on the photodetector performance. The Rλ and D* of
FeS2 devices at 1000 nm illumination show 0.16 A/w and 5.25
× 1010 Jones (10 min aging), 0.60 A/w and 8.41 ×
1010 Jones (40 min aging) and 3.94 A/w and 1.16 ×
1011 Jones (240 min aging). The enhanced detectivity can be
attributed to the increased crystallinity as a result of increased aging time during the
FeS2 growth. These results support that defects within the material are being
mitigated due to the recrystallization of the FeS2 nanomaterials. More detailed
photodetector device studies are underway to utilize their unique IR absorbance.
Figure 6
The FeS2 nanosheet photodetector performance.
(a) The reproducible on/off switching of the device upon AM-1.5 sun light at a bias of
1.0 V. (b) Detectivity under 500 nm and 1000 nm light illumination,
dependent on the growth time of FeS2 sheets (inset shows the schematic of
FeS2 sheets device).
A symmetry-defying OA growth and its implications for different shaped FeS2
nanocrystals have been presented and discussed. FeS2 nanocrystals show the growth
starting with FeS2 QD seeds, which exhibit excitonic absorption behavior and
enable further OA growth for shape and size control. A growth pathway model and
thermodynamic reasoning are then presented to facilitate understanding of shape and size
control in the FeS2 system. Shape and crystallinity of FeS2
nanocrystals is shown to be dependent on reaction temperature and aging time. Photodetector
performance is shown to be correlated with crystallinity, offering support for defect
mitigation in the material. Observation of the symmetry-defying OA growth in FeS2
nanocrystals and its effect on crystallinity will facilitate FeS2 along on its
path to becoming a “golden” material for sustainable energy applications.
Controlling crystallinity is a key point in the generation of complex functional
nanomaterials. Self-assembly of particles into larger single-crystalline objects by the OA
mechanism, is one of the most promising approaches in nanotechnology. This OA evolution
process can be adjusted by cosolvents2631, high pH value34,
temperature and time24. A well-controlled reaction conditions in the OA
process can facilitate the high quality nanocrystal growth.
Method
Synthesis
The FeS2 nanocube synthesis starts with 0.5 mmol FeCl2 in
octadecylamine (ODA, 12 g) loaded into a three neck flask and degassed and back
filled with argon, heated to 393 K and allowed to decompose for 120 min.
Another three neck flask is then loaded with 4 mmol sulfur powder in diphenyl ether
(5 mL), is degassed and back filled with argon, and heated to 343 K for 1
hour to dissolve. The sulfur solution is then quickly injected into the Fe-ODA precursor
at the temperature of 393 K. After injection, the combined solution was heated to
493 K and aliquots at different time intervals were taken for UV-vis-NIR absorption
test and HRTEM characterization. For FeS2 thin sheets, injection temperature of
the Fe-ODA precursor is raised to 418 K with everything else kept the same.
Particles were separated by centrifugation and purified by being re-dissolved and crashed
in chloroform-methanol. The final particles were dispersed in chloroform for storage and
characterization.
Materials Characterization and Devices fabrication
All UV-Vis-NIR absorbance spectra were obtained on a UV-3600 Shimadzu Spectrophotometer.
HRTEM images were obtained using Field Emission FEI Tecnai F20 XT. The photodetector
devices are fabricated as following: A PEDOT:PSS layer is used to flatten the ITO
patterned glass substrate and serves as a hole transporting layer. The FeS2
nanosheets were dissolved in chloroform with a concentration of 25 mg/mL. The
FeS2 nanosheets were deposited on the PEDOT:PSS surface by spin coating
method at the speed of 1500 RPM, Then, a thin layer of calcium (~10 nm) was
thermally evaporated. Finally, a patterned aluminum electrode (~80 nm) was
evaporated on the top surface of the calcium, completing the device.
Author Contributions
M.G. and A.K. carried out the synthesis, characterization and analyzed the data, and wrote
the paper. The authors M.G. and A.K. contributed equally to this work. S.R. supervised the
project and conceived the idea and experiments. All the authors discussed the results,
commented on and revised the manuscript.