Continuous-flow hydrothermal

synthesis for the production of

rsta.royalsocietypublishing.org
inorganic nanomaterials
Peter W. Dunne, Alexis S. Munn, Chris L. Starkey,
Review Tom A. Huddle and Ed H. Lester
Cite this article: Dunne PW, Munn AS, Department of Chemical and Environmental Engineering, University
Starkey CL, Huddle TA, Lester EH. 2015 of Nottingham, University Park, Nottingham NG7 2RD, UK
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

Continuous-flow hydrothermal synthesis for

the production of inorganic nanomaterials. As nanotechnology becomes increasingly important
Phil. Trans. R. Soc. A 373: 20150015. and ubiquitous, new and scalable synthetic
http://dx.doi.org/10.1098/rsta.2015.0015 approaches are needed to meet the growing
demand for industrially viable routes to nanomaterial
production. Continuous-flow hydrothermal synthesis
Accepted: 29 June 2015
or supercritical water hydrothermal synthesis
(scWHS) is emerging as a versatile solution to
One contribution of 12 to a discussion meeting this problem. The process was initially developed
issue ‘Supercritical fluids: green solvents for to take advantage of the tunable chemical and
physical properties of superheated water to produce
green chemistry?’
metal oxide nanoparticles by rapid nucleation and
precipitation. The development of new mixing
Subject Areas:
regimes and reactor designs has been facilitated by
chemical engineering, nanotechnology, the modelling of reactor systems. These new reactor
materials science designs further exploit the properties of supercritical
water to promote faster and more uniform mixing
Keywords: of reagent streams. The synthetic approach has
continuous-flow, supercritical water, been expanded beyond the metal oxide systems for
reactor design, nanomaterials, metal–organic which it was conceived, and now encompasses metal
sulfides, metal phosphates, metal nanoparticles and
frameworks
metal–organic frameworks. In many of these cases,
some degree of size and shape control can be achieved
Author for correspondence: through careful consideration of both chemistry and
Ed H. Lester reactor design. This review briefly considers the
e-mail: [email protected] development of scWHS reactor technology, before
highlighting some of our recent work in expanding
the scope of this synthetic method to include a wide
range of materials.

1. Introduction
The synthesis of inorganic nanomaterials is driven by
the ever-increasing array of applications found to benefit
from the reduction in size of the primary particles.

2015 The Author(s) Published by the Royal Society. All rights reserved.
 These benefits can arise simply from the fact that nanoparticles may, through a combination of
2
their small size and adjustable surface chemistry, be made highly dispersible or soluble [1–4],
allowing simple solution-based processing for the formation of coatings, thin films or dense

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
structured ceramics. Moreover, materials confined to the nanoscale often display properties vastly
different from those of their bulk counterparts. The reduction in size may lead to changes in
structure and an increase in surface defects, as demonstrated in the case of ceria [5,6]. Small
particle sizes also lead to increased surface areas, and these two factors combine to offer enhanced
catalytic activity. When the particles are sufficiently small, many materials exhibit quantum
confinement effects, causing a change in the optoelectronic properties of the material [7–9].
Material properties can also vary with varying size and shape, enabling the development of
dye-sensitized solar cells [10] and quantum dots, which find use in biomedical imaging [11]
and display devices [12]. The enhanced physical, chemical and optoelectronic properties of
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

many nanomaterials have thus led to them finding use in a vast array of technological
applications, including catalysis [13], photocatalysis [14], solar cells [15], batteries [16], light-
emitting diodes [17], gas sensors [18] and more. As the demand for nanomaterials increases, new
methods for their synthesis are constantly being developed.
Traditionally, inorganic materials are synthesized by high-temperature ceramic methods [19].
The high temperatures, typically in excess of 1000◦ C, prohibit chemical flexibility and size
or shape control. Physical/thermochemical synthetic techniques, such as physical vapour
deposition and chemical vapour deposition [20–22], and other spray pyrolysis methods have
long been favoured in the industrial production of nanomaterials, as they are typically cheap
and relatively easy; however, they can suffer from high capital costs and the frequent need for
toxic precursors. These techniques are largely restricted to the production of coatings and films.
Solution-phase methods to produce inorganic nanomaterials have become increasingly common,
as a result of the much greater chemical flexibility and potential for size and shape control. Sol–gel
methods, involving the controlled hydrolysis of metal salts or alkoxides, followed by thermal
treatment of the resulting gels, have been used extensively for many years in the synthesis of
metal oxide nanoparticles [23]. Sol–gel methods typically yield highly homogeneous samples,
though the thermal treatment step may cause unwanted sintering. Micelle-mediated reactions
have also found favour for the ability to effectively control the size and porosity of the produced
nanoparticles by the use of self-assembled micellar nano- or micro-reactors in which metal
cations and precipitating agents are brought together to yield well-defined nanoparticles [24].
Unfortunately, the complexity of these systems makes them relatively low-yield and
high-cost processes.
Hydrothermal and solvothermal synthesis techniques, whereby precursor solutions are heated
in a sealed vessel at temperatures above the boiling point of the solvent, allow an enormous
degree of freedom in terms of the available chemistry [25]. The choice of solvent, pH, reagents
and additives invites the design of elegant reaction systems. The moderate temperatures at which
these reactions take place also permit the synthesis of metastable phases that may not be accessible
by other means [26]. While these conditions are relatively mild when compared with ceramic
methods, it is often the case that the products obtained are highly (nano-)crystalline in nature,
without the need for further heating steps. Hydrothermal and solvothermal syntheses, however,
tend to require long reaction times. While significant progress has been made in this area, there
remains little awareness of the physical conditions within the reaction vessels. Potentially, uneven
heating and prolonged reaction times often lead to complications such as phase separation
and polydispersity. As with the aforementioned sol–gel and micelle techniques, hydrothermal
synthesis is most typically a batch process, and, as with all batch processes, is not easily scalable.
As industrial demand for nanomaterials increases, the need for scalable, cost-effective and flexible
methods of production becomes ever more pressing.
Over the past 20 years, continuous-flow hydrothermal and solvothermal synthesis methods
have been investigated as highly promising one-step processes for the production of inorganic
nanomaterials. They have the potential to provide a wide variety of nanoparticles cheaply
and effectively at the scales required to supply the increasing industrial demand. This review
 initially examines the factors that must be considered in the design and implementation of
3
continuous-flow reactors for the controlled production of inorganic nanomaterials, and how
reactor design has evolved to meet these challenges, before considering some of the materials

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
produced by continuous-flow hydrothermal and solvothermal methods.

2. General considerations
(a) Nucleation and growth
There are many factors to be considered in the synthesis of nanoparticles. The ideal for most
purposes is to obtain nanoparticles with controlled sizes and a narrow size distribution. The
LaMer model (figure 1) for the growth of monodisperse colloids, developed in the 1950s for
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

colloidal sulfur, offers a simple conceptual framework with which to consider the formation
of nanoparticles [27]. Upon mixing, or otherwise initiating a reaction, there is a build-up
of precursors (stage I; figure 1), creating a degree of supersaturation. When the level of
supersaturation surpasses the critical nucleation threshold, nucleation occurs (II), reducing the
degree of supersaturation. If the rate of nucleation outstrips the rate of precursor formation,
the precursor concentration will drop back below the nucleation threshold. Nanoparticle growth
may then proceed by a number of pathways: monomer addition, in which additional precursor
units deposit onto the preformed nuclei from solution; Ostwald ripening, whereby energetically
disfavoured small nuclei redissolve and deposit onto more thermodynamically favourable
larger nuclei; and coalescence, in which multiple nanoparticles come together and fuse (III).
This framework allows us to consider methods by which both size and size distribution may
be controlled.
In most batch processes, such as sol–gel or conventional hydrothermal and solvothermal
reactions, precursor formation occurs over a long time period owing to, for example, slow heating
rates. As a result of this, the precursor concentration is often maintained above the critical
nucleation threshold for extended periods, and nucleation occurs continuously and concurrently
with growth. This gives rise to broad particle size distributions. If sufficiently long reaction
times are employed, then particle size will ‘focus’. Focusing occurs when nuclei formed over the
course of the reaction grow by addition from solution, giving initially a polydisperse sample. As
suggested by Sugimoto [28], the rate of diffusion-limited growth will vary with particle radius
in such a way that small particles will grow faster than larger ones. Coupled with Ostwald
ripening and coalescence, this leads to the focusing effect—a narrowing of the size distribution,
with the formation of larger particles at the expense of smaller, less stable nuclei. Thus, while the
particle size distribution may narrow, the actual size increases significantly. This is, in effect, the
rationale behind many sol–gel and hydrothermal/solvothermal batch processes, which typically
yield quite large nanoparticles. Carefully considered chemistry in the choice of solvents, reagents
and additives such as capping agents and micelles to control the kinetics of both nucleation and
growth can mitigate these size increases while maintaining a narrow size distribution, though this
adds further complications and the need for long reaction times remains.
An alternative strategy, which can potentially decrease the required reaction times
significantly, is the temporal separation of the nucleation and growth stages. In this case,
nucleation must effectively happen only once, through a burst nucleation step. This can be
achieved by the extremely fast generation of precursor such that the critical nucleation threshold
is passed quickly. Rapid nucleation then relieves the supersaturation. This initial nucleation
step must be sufficient to bring the precursor concentration back below the critical point, and
without the addition of further precursor, no new nuclei may be formed. By achieving this
fast burst of nucleation, there are minimal size discrepancies, thus limiting Ostwald ripening.
This leads to nuclei of similar size, which grow almost exclusively by diffusion. This gives
a narrow size distribution, with the possibility of size control by varying reaction times and
the addition of further precursors while maintaining the concentration below that required for
nucleation to occur. This approach has been used extensively in hot-injection processes for the
 4
Climit limiting supersaturation

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
Cnucleation nucleation

concentration
diffusion growth
Csolubility
Ostwald ripening

I II III
time
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

Figure 1. The LaMer model of nucleation and growth. (Online version in colour.)

growth of monodisperse metal, metal oxide and chalcogenide nanoparticles, particularly in the
production of quantum dots, where small sizes and narrow size distributions are critical [29,30].
To prevent growth and to limit the nanoparticle size to near that of the initially formed nuclei, it
is possible to induce the nucleation burst and then immediately stop any further reaction. This
can be achieved by limiting the available precursor, either by ensuring that the initial nucleation
consumes all available precursor, by quenching the reaction mixture or by removing any driving
force necessary for further growth, e.g. temperature.

(b) Properties of supercritical water and reactor design

Many of the reactions that generate inorganic nanomaterials require organic solvents, such
as alcohols for sol–gel syntheses, or high-boiling-point coordinating solvents for hot-injection
methods. Water is, of course, the ‘greenest’ solvent available. However, its physical and chemical
properties often prohibit its use in such systems. It has a relatively low boiling point that limits
its applicability to many high-temperature reactions and has a high polarity that prevents the
solubilization of many organic and non-polar molecules that may be needed as capping agents
to stabilize inorganic nanomaterials. These limitations can of course be overcome. Temperatures
above 100◦ C can be easily achieved for water when pressure is applied, whether external or
autogenous. Under such conditions, particularly on approaching the critical point of 374◦ C and
22.4 MPa, water exhibits many unusual properties, as may be expected of such a complicated
‘simple’ molecule. Such high temperatures permit reactions that would not otherwise be possible.
At high temperatures, the density and polarity of water are vastly decreased, and this is more
pronounced upon reaching the critical point [31,32]. Furthermore, the dissociative behaviour
changes drastically with temperature. These factors (highlighted in figure 2) have enormous
implications on reactor design.
The simplest conceivable reactor design for a continuous-flow hydrothermal/solvothermal
process is one in which a precursor solution is simply pumped through a heated zone under
pressure, inducing reaction, with the products being collected after passing through the heated
zone and any cooling stages required. While this approach has been used by a number of
researchers, particularly in microfluidic reactors, and has proven successful in the synthesis of a
variety of materials, it does not take advantage of the possibilities offered by reactor engineering.
The effect of this design on nucleation and growth is similar to that observed in a standard batch
hydrothermal reaction, with new nuclei continuing to form while previously formed nuclei grow
as the reactant stream flows through the reactor. However, unlike traditional batch hydrothermal
methods, reaction times are limited by the heated volume of the reactor and the necessary flow
rates. As such, it is difficult to achieve sufficiently long reaction times in which all nuclei are under
the same conditions for enough time to promote the focusing effect to control the size distribution.
 (a) (b)
5
supercritical –10
fluid 1.0 80
–12

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
(374°C,
220 bar)
0.8 60
–14

dielectric constant
density (g cm–3)
pressure (bar)

–16
0.6

logKw
40
solid liquid –18
0.4
–20
20
–22
(100°C, 0.2
1.01 bar)
(0.01°C, density –24 0
0.006 bar)
vapour logKw
0
dielectric constant –26
–20
–200 –100 0 100 200 300 400 500 0 50 100 150 200 250 300 350 400 450 500
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

temperature (°C) temperature (°C)

Figure 2. A (very) simplified phase diagram of water (a), and selected properties of supercritical water (b). (Online version in
colour.)

(a) (c)

(b)

Figure 3. Results of pseudo-fluid modelling studies on the mixing dynamics of different reactor geometries: T-piece (a), Y-mixer
(b) and co-current (c). Red represents the heated flow, and blue the cold stream. (Online version in colour.)

The products from this basic system suffer from poor size control, as subsequent growth of nuclei
is an unavoidable consequence of the simple design; consequently, it produces materials with a
high degree of polydispersity.
In 1992, Adschiri et al. [33] exploited the properties of near- and supercritical water, namely the
decreased dielectric constant and increased dissociation to H+ and OH− , to facilitate the rapid
continuous production of a variety of metal oxides in what has become known as supercritical
water hydrothermal synthesis (scWHS) and which is more generally known as continuous-flow
hydrothermal synthesis (CFHS). Here a reagent stream, typically an inorganic metal salt (MLx ), is
brought into contact with a separate stream of water pre-heated to a near-critical or supercritical
state. The high concentration of OH− species in near-critical water leads to the immediate
 6

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
water (pre-)heater
HPLC pump
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

cooler
mixing point

back pressure
regulator

precursor 1 HPLC pump product collection

HPLC pump
precursor 2

Figure 4. A schematic diagram of the Nottingham counter-current reactor. The inset shows the mixing regime within the
reactor. (Online version in colour.)

hydrolysis and dehydration of the metal salt to generate the metal oxide, which immediately
precipitates as nanoparticles, according to the equations

MLx + xH2 O → M(OH)x + xHL

M(OH)x → MOx/2 + (x/2)H2 O.

These metal oxide nanoparticles are then carried further downstream through an additional
heating zone to promote growth. This process essentially replicates the batch hot-injection
procedure in a continuous-flow system, and the basic principle now informs the majority of
continuous-flow hydrothermal techniques.
Early work on scWHS reactors used simple T-piece mixing points. The mixing efficiency
within these reactors has been studied using pseudo-fluid modelling, whereby methanol and
sugar water, with densities of 0.79 and 1.178 g cm−3 , respectively, are used as analogues for
low-density supercritical water and the higher-density cold stream [34]. These results, shown
in figure 3a, indicate that T-piece mixers do not provide uniform mixing and suffer from
undesirable premixing of the two streams. Computational studies to model the mixing regime
within T-piece reactors have also shown that there are disadvantages with this type of reactor
owing to the incomplete mixing of reagent streams. Alternative reactor arrangements have also
 been examined, including Y-mixing and co-current mixing (figure 3b,c). Through these studies,
7
design criteria were identified for an ideal scWHS reactor:

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
(1) Instant, uniform and complete mixing of the reagent streams to yield a high number of
small nuclei.
(2) Short average residence times and narrow residence time distributions in order to
maintain small size and narrow size distributions, respectively.
(3) Minimal heating of the precursor stream prior to mixing to avoid premature nucleation,
unwanted reactions and blockage of the tubing prior to the reactor.
(4) Adequate flow to prevent particle settling and accumulation within the reactor.

The application of these criteria led to the development of the counter-current mixing reactor
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

used in our laboratories in Nottingham [35]. A supercritical water stream is achieved by

pumping water through a pre-heater unit before passing downwards through a central inner
tube to a mixing point, where it meets a room-temperature reagent stream flowing upwards.
This vertical tube-in-tube arrangement exploits the lower density of the supercritical stream to
promote rapid and symmetrical mixing, whereas the physical arrangement prevents undesirable
mixing or heating of the reagent stream. Both pseudo-fluid modelling and computational studies
suggest that this design overcomes many of the problems associated with the more basic T-piece
reactors. A schematic of this reactor and pseudo-fluid modelling, highlighting the efficient mixing
dynamics, is shown in figure 4. Unless otherwise stated, this is the reactor configuration used in
the work reviewed here.

3. Materials
As mentioned above, the scWHS method was initially developed for the production of fine metal
oxide nanoparticles, and much of the work in the field has remained focused on these materials.
Our own work includes many metal oxide systems, and we have also been involved in a concerted
effort to expand the library of materials accessible by this technique. Several reviews have been
written on the continuous-flow hydrothermal synthesis of metal oxide nanoparticles [36,37],
and as such we will only consider them here briefly. The remainder of this review provides an
overview of our work in mapping the synthetic landscape around this technology. An overview
of some of the materials prepared by continuous-flow hydrothermal and solvothermal techniques
is given in table 1.

(a) Metal oxides

The initial investigations, performed by Adschiri et al. using a T-piece mixer-type reactor, explored
the synthesis of a whole host of metal oxides, including haematite (Fe2 O3 ), magnetite (Fe3 O4 ),
cobalt oxide (Co3 O4 ) and anatase-phase titania (TiO2 ) [33]. Further work by Adschiri et al.
added ceria (CeO2 ) [38], lithium cobalt oxide (LiCoO2 ) [58] and the oxyhydroxide boehmite
(AlOOH) [59]. This approach informed much of the subsequent research in the area, and the
number of oxides available by this technique expanded to include more complex oxides such
as yttrium aluminium garnet (Y3 Al5 O12 ) [40,41], its doped analogues and potassium niobate
(KNbO3 ) [42]. Other mixed metal oxides, such as Ba(1−x) Srx TiO3 , have also been prepared via
solvothermal methods using alkoxide precursors with longer residence times [43,44]. As is shown
in figure 5 in the case of both titania and haematite, the size of the obtained nanoparticles may
be altered by changing the reaction temperature. Generally, the size increases with increasing
reaction temperature, indicating an increase in the critical nucleus size at higher temperatures.
While this is difficult to reconcile with the decreased polarity of supercritical water, it is
consistently observed across many (but not all) systems, and it may be the case that the increased
temperature overrides the decreased solubility. Alternatively, it may be that, in these particular
systems in which the oxide can be formed at temperatures as low as 200◦ C, higher temperatures
 Table 1. A selection of materials prepared by continuous-flow hydrothermal and solvothermal methods (∗ , this work).
8
material solvent temp. (◦ C) notes size (nm) shape reference

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
oxides TiO2 H2 O 400 20, 3–8∗ spherical [33], this work
.....................................................................................................................................................................................................................

CeO2 H2 O 400 varied times 20–300 octahedral [38]

.....................................................................................................................................................................................................................

ZrO2 H2 O 400 10 spherical [33]

.....................................................................................................................................................................................................................

Fe2 O3 H2 O 200–400 50, 5–30∗ spherical [33], this work

.....................................................................................................................................................................................................................

Fe3 O4 H2 O 50 spherical [33]

.....................................................................................................................................................................................................................

Co3 O4 H2 O 200–430 9–13 cubic [39]

.....................................................................................................................................................................................................................

Y5 Al3 O12 H2 O spherical [40,41]

Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

.....................................................................................................................................................................................................................

KNbO3 H2 O 400–410 30–50 spherical [42]

.....................................................................................................................................................................................................................

Ba(1−x) Srx TiO3 H2 O : EtOH 150–380 varied times 15–35 cuboidal [43,44]
.............................................................................................................................................................................................................................................
sulfides ZnS H2 O 250–400 nucleation 3–11 spherical [45]
..............................................................................................................................................................

400 growth 200 nanoflower [45]

.....................................................................................................................................................................................................................

CdS H2 O 250–400 nucleation 7–19 spherical [45]

..............................................................................................................................................................

400 growth 100 × 40 rods, multipods [45]

.....................................................................................................................................................................................................................

PbS H2 O 250–300 nucleation 15–27 cuboidal [45]

.....................................................................................................................................................................................................................

CuS H2 O 250 nucleation 40 × 24 hexagonal platelets [45]

.....................................................................................................................................................................................................................

Fe(1−x) S H2 O 400 nucleation 100 × 30 hexagonal platelets [45]

..............................................................................................................................................................

400 growth 1000 × 30 hexagonal plates [45]

.....................................................................................................................................................................................................................

Bi2 S3 H2 O <250 nucleation 300 × 20 tubes [45]

.....................................................................................................................................................................................................................

MoS2 H2 O 400 → 250 multi-step 1000 × 9 tangled sheets [46]

.............................................................................................................................................................................................................................................
phosphates LiFePO4 H2 O 400 extra heating various∗ various [47–49], this work
.....................................................................................................................................................................................................................

Ca5 (PO4 )3 OH H2 O 300 pH 8 250 × 100 plates [50]

..............................................................................................................................................................

300 pH 10 1000 × 50 rods [50]

..............................................................................................................................................................

400 rods [51,52]

..............................................................................................................................................................

400 pH 10 1000 × 50 tubes [50]

.............................................................................................................................................................................................................................................
metals Ag H2 O 400 various [53]
.......................................................................................................................................................................................

H2 O 400 PVP spherical [53]

.......................................................................................................................................................................................

MeOH 400 30 spherical [54,55]

.......................................................................................................................................................................................

EtOH 400 320 spherical [55]

.....................................................................................................................................................................................................................

Cu H2 O 350–400 formic acid, PVP 17–25 spherical [56]

.......................................................................................................................................................................................

MeOH 400 33 spherical [54,55]

......................................................................................................................................................................................................

Ni MeOH 400 30 spherical [54,55]

.............................................................................................................................................................................................................................................
MOFs CPO-27 DMF 200–300 12 spherical [57]
.......................................................................................................................................................................................

DMF 200 low conc. 200 × 50 needle [57]

.....................................................................................................................................................................................................................

HKUST-1 EtOH : DMF 300 5–10 000 truncated octahedra [57]

.....................................................................................................................................................................................................................

ZIF-8 H2 O 400 2000∗ truncated rhombic this work

dodecahedra
.......................................................................................................................................................................................

H2 O 100 4000∗ bevelled cube this work

.............................................................................................................................................................................................................................................
 allow nanoparticle growth to continue, as it takes longer for the system to cool. Typically, the
9
metal oxide nanoparticles produced by this method adopt a roughly spherical morphology,
adding credence to the proposed fast nucleation mechanism. There are exceptions to this, of

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
course, with spinel-type Co3 O4 , for example, adopting a highly faceted cubic morphology [39].
Clearly, given the temperature and time dependence of particle size, and the observation of
faceted nanostructures, it must be assumed that some degree of growth occurs even in this
highly nucleation-biased process. It is reasonable to suggest that, even at such high temperatures,
incomplete reaction at the mixing point allows for some diffusional growth before the flows can
be cooled below the temperatures required for further growth and crystallization.

(b) Metal sulfides

Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

During the synthesis of metal oxide nanomaterials, the oxygen is provided by the water itself.
Consequently, consideration must first be given to how a suitable sulfur source can be introduced
into the existing method that will promote metal sulfide production in preference to metal oxide
production. Many sulfur-containing compounds will break down under hydrothermal conditions
to hydrogen sulfide or hydrosulfide (H2 S or HS− ), but these are often highly volatile, toxic,
poorly water-soluble and have unpleasant odours (to say the least). While sodium sulfide is an
obvious contender, its dissociation in water limits its practical application, as precipitation of
poorly crystalline metal sulfides occurs immediately with little control. To this end, thiourea was
chosen as a suitable precursor. It is odourless, water-soluble and through the isomerization to
ammonium thiocyanate has a very low-energy route to generating hydrogen sulfide [60]. Once
produced, hydrogen sulfide may essentially act in place of hydroxide ions in the oxide production
scheme, thus providing a route to metal sulfide nanomaterials:

NH2 CSNH2 ↔ NH4 SCN

NH4 SCN + 4H2 O → HS− + (NH4 )2 CO3 − + H2 O
Mx+ + xHS− → MSx/2 + (x/2)H2 S.

An interesting complication arises from the in situ generation of HS− as a reagent in a flow
system such as this: there are multiple options for delivering the sulfur source. When the thiourea
is flowed through the pre-heater, the decomposition reaction will be complete prior to mixing
with the metal salt stream, thus ensuring an excess of hydrogen sulfide at the mixing point.
This route promotes a nucleation-driven reaction. Alternatively, the thiourea and metal source
may both be contained in the cold reagent stream, such that thiourea decomposition and metal
sulfide formation are in competition after mixing with the superheated water stream. This latter
route maintains the overall precursor concentration at lower levels and promotes growth of
the metal sulfide particles. Figure 6 shows the application of these different methods to the
production of various metal sulfides. Under the nucleation-driven method, small, typically sub-
20 nm, nanoparticles are obtained for many different sulfides, including zinc, cadmium and lead.
Because of the longer residence time afforded to thiourea decomposition using the nucleation
method, it has been found that this method is more widely applicable because it offers a wider
range of reaction temperatures. Specifically, lead sulfide, copper sulfide and bismuth sulfide may
only be prepared at subcritical temperatures, as reaction temperatures above 300◦ C generate
impurities or cause decomposition of the desired sulfide phases. The growth method in contrast
is only possible with the pre-heated water stream at supercritical conditions, as this burst of high
temperature on mixing is required to initiate the breakdown of thiourea in the short contact time.
As this breakdown happens concurrently with metal sulfide formation, larger nanostructures of
zinc, cadmium and iron sulfide are obtained. The final morphologies are heavily predicated by
the underlying crystal structure. A selection of the metal sulfide nanoparticles and nanostructures
derived from these methods are shown in figure 6 [45].
Given the wide applicability of the nucleation method for the production of metal sulfide
nanoparticles, we sought to apply this same technique to the generation of molybdenum
 (a) (b)
10
10
9

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
425°C 8

crystallite diameter (nm)

7
6
375°C
5
4
325°C
3
275°C 2
1 titanium oxysulphate
225°C
0
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

20 30 40 50 60 200 250 300 350 400 450

2q (deg.) temperature (°C)

(c) (d) (e) (f)

100 nm 100 nm 50 nm
50 nm

(g) (h)

50 nm 10 nm

(i) ( j)

50 nm
100 nm

Figure 5. X-ray diffraction patterns and crystallite sizes of anatase TiO2 nanoparticles obtained at different reaction
temperatures (a,b). TEM images of: haematite nanoparticles produced at 250◦ C, 300◦ C, 350◦ C and 400◦ C (c–f , respectively);
zirconia produced under supercritical conditions (g,h); cobalt oxide nanocubes (i); and YAG nanoparticles ( j). (Online version in
colour.)

sulfide, an important layered chalcogenide with applications in high-temperature lubricants,

catalysis and electronics. It was found that, starting from the widely available Mo(VI) precursor
ammonium molybdate, only the thiomolybdate anions MoO(4−x) Sx 2− could be obtained, with an
increasing degree of sulfidation on increasing the reaction temperature [61]. Under supercritical
 nucleation growth
11
mixing nucleation method
point growth method
CdS

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
nucleation

[Mx*] + [HS]
DG
r*
radius T1 T2 T3 T4 C*

growth

no reaction Fe(1–x)S

time/position

Bi2S3 Fe(1–x)S CuS PbS CdS ZnS ZnS

Figure 6. A range of metal sulfide nanostructures produced by nucleation- and growth-driven processes in the counter-current
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

reactor. (Online version in colour.)

conditions, the tetrathiomolybdate anion MoS4 2− was formed. This readily transforms into
amorphous Mo(IV)-containing MoS3 under acidic conditions [62]. Acidic conditions could
be achieved easily in flow simply by the addition of a further inlet stream after the initial
mixing point. The acid chosen for this is acetic acid, as it is non-oxidizing to promote the
Mo(VI) → Mo(IV) reduction while also being compatible with the stainless steel reactor at high
temperatures. The modular nature of the continuous-flow system also allows an additional
heating unit to be added to the outlet of this reactor to promote the crystallization of MoS2 from
the amorphous MoS3 precursor generated in the preceding steps [46]. This multi-step reaction
produces tangled nanosheets of MoS2 , as shown in figure 7, and highlights the value of both a
continuous-flow system and supercritical water in taking a reaction that requires days to perform
in batch to completion within approximately 30 s:

MoO4 2− + 4HS− → MoS4 + 4OH−

MoS4 2− + H+ → MoS3 + HS−

MoS3 → MoS2 .

(c) Metal phosphates

Perhaps the biggest driving force behind the exploration of CFHS for the production of
phosphate materials lies in the emergence of lithium iron phosphate as the next-generation
battery material [63]. The ever-increasing push towards green and sustainable energy with the
need for large-scale grid storage, as well as the burgeoning abundance of electric vehicles and
personal electronic devices where miniaturization is key, means that new and improved battery
materials are in constant demand. Lithium iron phosphate emerged as a contender for the throne
of ‘next-generation battery material’ some years ago, following on from its initial production
and the speculations of Padhi & Goodenough [64] regarding its potential use in electrochemical
devices. With its high capacity, stability and low toxicity, LiFePO4 offers many advantages
over conventional metal oxide-based cathode materials (LiMO2 ). However, it suffers from low
electronic conductivity and sluggish lithium-ion diffusivity through the olivine structure. These
shortcomings may be mitigated or overcome by controlling the size and shape of the LiFePO4
in order to minimize the Li+ diffusion path length [65]. CFHS offers a route to large-scale
synthesis, sufficient to meet the growing demands of a society that is increasingly dependent
on electricity, with the scope to produce small nanostructures with controlled shapes. A number
of groups have been active in this area, and a great wealth of work now exists exploring
the ramifications of reaction parameters on the production of LiFePO4 [47–49]. It has been
found by these researchers that, under subcritical conditions, many impurity phases may be
 thiourea
0.1 M 12
20 ml min–1

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
pre-heater
T: 400°C

2
0.2 mm

acetic acid
3
20% (v/v)
10 ml min–1
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

MoOS32– cooling
MoO3S2– product
1 and
collection

intensity (arb. units)

3

MoO2S22– MoS42–

ammonium heptamolybdate
0.05 M
10 ml min–1 10 20 30 40 50 60 70
2q (deg.)

Figure 7. The sequential continuous-flow synthesis of molybdenum disulfide. (Online version in colour.)

obtained. In a typical continuous hydrothermal reaction, whereby a mixed solution of H3 PO4

and FeSO4 is mixed with LiOH prior to meeting a pre-heated water stream, oxidation of Fe(II)
to Fe(III) occurs, resulting in iron oxide impurities. Similarly, the widely accepted mechanism
for LiFePO4 formation involves the initial production of iron phosphates (Fex (PO4 )y · zH2 O),
and such materials are often observed as impurities. Under supercritical conditions, many of
these issues are resolved (figure 8), particularly the presence of iron phosphate impurities.
Preventing the formation of iron oxides remains a challenge, even at supercritical conditions,
and is an important barrier to overcome, because impurity phases are detrimental to the eventual
performance of the LiFePO4 cathode, although careful degassing of precursor solutions can help
with this. Despite a number of issues remaining with continuous hydrothermal synthesis of
LiFePO4 , large-scale production is being realized, and is fast becoming a highly competitive
and lucrative market. In 2011, the Korean company Hanwha Chemicals completed the world’s
first continuous supercritical hydrothermal plant for the production of nanoscale lithium iron
phosphate, producing at 100 tonne per year scale (http://hcc.hanwha.com/eng/index_eng.jsp),
clearly demonstrating the capabilities of continuous-flow supercritical hydrothermal synthesis
as an industrial process. With an intense push recently by automobile manufacturers and
governments worldwide towards the electrification of vehicles, this need for ever cheaper and
sustainable synthesis technology is only going to grow more rapidly.
In addition to the extensively studied lithium iron phosphate, the CFHS method has been
investigated as a route to the calcium phosphate mineral hydroxyapatite, Ca5 (PO4 )3 OH, which is
widely used in biomedical applications. Given its natural role in the body as a component of teeth
and bones, synthetic hydroxyapatite is ideally suited as a bioceramic, acting as a replacement
for damaged bone, a coating to promote regrowth and as a porous material for drug delivery
in bone implants. Given the obvious benefits of the nanoscale-based applications and the future
scope for three-dimensional printing of bioceramics, it is likely that methods for the production of
biocompatible nanoceramics will be highly sought after. Several research groups have reported on
 pre-heater
water
T: 400°C
13

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
additional heating

product cooling
and
collection 200 nm
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

FeSO4
H3PO4 LiOH

Figure 8. The supercritical water hydrothermal synthesis of lithium iron phosphate nanoparticles. (Online version in colour.)

pH 8 pH 10 pH 10
300°C 300°C 400°C

50 nm 100 nm 200 nm

Figure 9. Hydroxyapatite nanoplates, nanorods and nanotubes synthesized in continuous flow under different conditions.

the production of hydroxyapatite nanostructures via the continuous-flow hydrothermal method.

Chaudhry et al. [51] first reported the synthesis of hydroxyapatite with a counter-current mixing-
type reactor in 2006. A calcium source is mixed in flow with a phosphate source prior to meeting
the pre-heated water stream that induces crystallization of the hydroxyapatite, where significant
increases in crystallinity upon increasing the temperature of the pre-heated stream were seen. The
best product from this method presents as low-aspect-ratio nanorods. Further work by Darr and
co-workers [52] developed routes to doped and substituted hydroxyapatites. Our own efforts in
this area have examined the scope for shape control within this system, with pH and reaction
temperature proving to be vital factors [50]. Figure 9 shows nanoplates, nanorods and nanotubes
of hydroxyapatite prepared under various conditions in our reactor system.

(d) Metals
Metal nanoparticles are subject to intense scrutiny not only owing to their importance as
model systems for nucleation and growth, but also because of their vast importance in many
technological fields. Biomedical imaging is assisted by the size- and shape-dependent optical
properties of colloidal metal nanoparticles [66], including gold and silver, whereas other precious
metal nanoparticles such as platinum, palladium and ruthenium serve as highly efficient
catalysts [67] in hydrogenation reactions and the removal of carbon monoxide, NOx and SOx
 from automotive exhausts, for example. More recently, there has been a growing interest in the
14
development of dispersible metal nanoparticles for the production of printable electronics [68,69],
the so-called electronic inks, with copper and silver proving most suitable for these purposes.

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
Certain metals may be prepared readily as nanoparticles by batch aqueous processes, gold and
silver being prime examples; however, many metal complexes are far too resistant to reduction.
In most cases, of course, a reducing agent is needed to generate metal nanoparticles from metal
salts. The citrate method for the controlled synthesis of colloidal gold is perhaps the most
widely known, though many organic and inorganic reducing agents have been used. This general
approach is easily transferred to a continuous-flow system. The production of silver nanoparticles
has been achieved on our system using an aqueous solution of silver acetate as the metal
source (figure 10a, b) [53]. The acetate moieties are believed to generate a sufficiently reducing
environment to cause the reduction to silver metal, though the products obtained are large and
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

highly polydisperse in size and shape. Additionally, the temperatures required for the reaction
are high enough to melt the silver nanoparticles, causing a build-up on the reactor walls. The
inclusion of polyvinylpyrrolidone (PVP) in the reagent stream as both a reductant and stabilizing
agent yields smaller and more uniform particles and prevents coalescence and deposition of the
silver nanoparticles (figure 10c, d). This reaction is possible owing to the ease with which silver(I)
may be reduced to silver metal (E0 = 0.8 V). In an oxygen-rich environment, such as is provided
by supercritical water, metal ions with lower redox potentials will tend to form the oxides. For
example, copper(II), which, nominally, is quite easily reduced to copper(0) (E0 = 0.34 V), cannot
reliably be reduced to produce copper nanoparticles under the same hydrothermal conditions,
generating copper oxide instead. In order to produce copper nanoparticles on a continuous-flow
system, it is necessary to perform the reaction under solvothermal conditions. Figure 10e shows
sub-10 nm copper nanoparticles obtained through a continuous-flow solvothermal reaction. In
this case, the organic solvent and its decomposition products are powerful enough reducing
agents to generate copper(0), as has been shown by Kim et al. [54,55] for the production of metal
nanoparticles in supercritical alcohols. Kubota et al. [56] have recently reported the continuous
hydrothermal production of PVP-stabilized copper nanoparticles using copper formate as the
copper source. Formic acid is well known to decompose to CO2 and H2 under hydrothermal
conditions, thus providing the impetus for the reduction reaction even in supercritical
water [70].

(e) Metal–organic frameworks

Metal–organic frameworks (MOFs) have been heralded in recent years as having the potential
to be industry-leading materials for various applications such as CO2 capture/storage, H2
storage, catalysis and pharmaceutical delivery systems. All of these potential applications involve
exploiting the porosity of the frameworks, which has been the main attraction of this group
of materials.
MOFs consist of two main components, an inorganic unit and an organic linker. The inorganic
unit can be either a single metal ion or a ‘metal cluster’, which normally comprises several metal
ions coordinated to a central electronegative atom. The metal ions or clusters are covalently
bonded to organic molecules; the ability to coordinate with these inorganic units is the main
restriction on the nature of this organic linker. The versatility of the organic unit is what has led
to the wide range of frameworks that have now been reported [71].
Conventionally, MOFs are synthesized using hydrothermal and solvothermal batch methods.
However, in recent years, novel methods for synthesizing MOFs have been reported, such as
microwaves [72–75], ultrasound [76] and mechanochemical synthesis [77]. These techniques can
benefit from reduced synthesis times and easier post-processing of products; however, none of
these techniques have been adapted for the continuous production of MOFs. There are only a
few reports for the continuous production of MOFs. Schoenecker et al. [78] in 2013 reported
a continuous process for the synthesis of UiO-66 using their patented reactor that uses batch
technology with an integrated flow system to allow for the removal of products and introduction
 50 nm (a) 50 nm (b) 15

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
(c) (d)
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

100 nm 100 nm

(e)

10 nm

Figure 10. Silver nanoparticles prepared without PVP at low and high concentrations (a,b), and with PVP (c,d), and copper
nanoparticles produced under continuous-flow solvothermal conditions (e).

of reagents in situ. The Serre research group described a scalable aerosol system, which uses
two liquid reagent streams injected into a flow of hot air, for the synthesis of HKUST-1, ZIF-8
and Fe3 (BTC)2 [79]. A similar spray-drying technique was also demonstrated by Carné-Sánchez
et al. [80]. A microfluidic system, demonstrated by Faustini et al. [81], was shown to be a versatile
technique for the production of various MOFs, including HKUST-1 and UiO-66, and several other
groups have investigated MOF synthesis in microfluidic systems [82,83].
To realize the industrial potential for MOFs, a continuous large-scale method for their
production needs to be developed. The continuous-flow reactor technology described above
was shown to be a viable option to solve this problem. In 2012, the successful synthesis
of two MOFs, HKUST-1 and CPO-27(Ni), was reported using this continuous method [57].
It was the first example of continuous hydrothermal and solvothermal synthesis of MOFs.
However, similar technologies have been reported since. Rubio-Martinez et al. [84] reported
the continuous synthesis of HKUST-1, UiO-66 and NOTT-400 using a reactor with a
T-piece mixer, which is followed by a heating unit, where the whole system is held under
a pressure of approximately 7 bar. A similar T-piece reactor with post-heating was used to
continuously produce a cerium–terephthalate MOF at 100 bar [85], and MIL-53(Al) at 250–300◦ C
 I
16

5 mm

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
I

(a) (b) 10 mm (c)

20 nm

5 mm
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

(d ) 100 nm (e) 10 mm ( f)

Figure 11. Electron microscopy images of MOFs synthesized at high temperatures and/or high concentrations (a) CPO-27,
(b) HKUST-1, (c) ZIF-8; and products synthesized at low temperatures and/or concentrations (d) CPO-27, (e) HKUST-1 and
(f ) ZIF-8.

and 230 bar [86]. A low-temperature (120◦ C), long-residence-time (60 min) flow system has also
been shown to produce a variety of MOFs by Waitschat et al. [87].
The Nottingham reactor has been used to make MOFs under a wide range of conditions from
room temperature to 400◦ C, from ambient pressure to 240 bar and using varying flow rates of the
available reagent streams. The work by the Stock research group shows that the choice of reagents
and the resulting pH levels can be critical in obtaining the desired product [88]. Our work has
shown that, by manipulating the synthesis conditions, the size and morphology of particles can
be controlled; particles have been produced on the micro- and nanoscale (figure 11). Early work
demonstrated that the method used for metal oxide synthesis, whereby all synthesis reagents
are introduced via the cold streams, could be directly applied to the synthesis of MOFs. It is
possible to synthesize MOFs under supercritical conditions; however, the method offers very little
advantage and introduces metal oxide impurities into the final product.

4. Conclusion and future outlook

This review highlights the extension of the supercritical water hydrothermal synthesis technique
beyond the production of fine metal oxide nanoparticles into a more widely applicable
continuous-flow hydrothermal and solvothermal route with scope for the large-scale production
of a wide variety of inorganic nanomaterials. Through a combination of reactor design and
‘chemical know-how’, a whole host of materials, including oxides, sulfides, phosphates, metals
and MOFs, are now available through this technique. It should be noted that adapting and
developing new synthetic strategies for these high-temperature, high-pressure continuous-flow
systems remains an empirical procedure. One of the great challenges remaining in this area
lies in understanding the critical physical and chemical processes around the mixing point. The
major difficulty here arises from the very nature of the technique, with the temperatures and
pressures prohibiting the monitoring of reactions by most conventional means, and thus limiting
the development of mechanistic theories to post-synthetic rationalizations. In situ monitoring of
these reactions by X-ray or neutron imaging and diffraction using synchrotron sources would
provide many answers to questions of reactor design and reaction mechanism. Iversen and
 co-workers [89,90] have made great strides in this area, with their work on the mechanisms
17
of crystallization in various systems providing insights into nucleation and growth phenomena
under supercritical conditions. The physical (and safety) constraints of working with supercritical

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
water have restricted these studies to post-mixing isothermal zones constructed of sapphire to
minimize attenuation of the X-ray beam. Attempts at directly observing the immediate mixing
point have often revealed only the build-up of product, which, while useful in its own right,
reveals little of the true chemistry occurring at this point [91]. With the continuing development
of synchrotron capabilities, it is expected that future experiments in this vein may prove more
fruitful. For now, the more old-fashioned empirical chemical methods have allowed enormous
progress, and this will continue, with new materials and further fine chemical and physical
control still to be achieved. Perhaps most importantly for this field, large-scale industrial facilities,
beyond Hanwha’s existing plant (http://hcc.hanwha.com/eng/index_eng.jsp), are being built
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

and will further prove the value of the technology, particularly as an industrial route to inorganic
nanomaterials.
Authors’ contributions. P.W.D., A.S.M. and C.L.S. drafted the paper and carried out the work on sulfides, MOFs
and LiFePO4 , respectively. T.A.H. carried out the pseudo-fluid modelling work.
Competing interests. We declare we have no competing interests.
Funding. This work is funded by the European Union’s Seventh Framework Programme (FP7/2007–2013),
grant agreement no. FP7-NMP4-LA-2012-280983, SHYMAN.
Acknowledgements. The authors thank Pablo Caramazana for results pertaining to titania and Dr Miquel
Gimeno-Fabra regarding the early MOF syntheses. We also thank Dr Sean Butterworth for the copper work,
and Promethean Particles Ltd. for access to data. The Nottingham Nanotechnology and Nanoscience Centre
and the Department of Mechanical, Materials and Manufacturing Engineering are gratefully acknowledged
for access to TEM and XRD facilities.

References
1. Peters K, Zeller P, Stefanic G, Skoromets V, Němec H, Kužel P, Fattakhova-Rohlfing D.
2015 Water-dispersible small monodisperse electrically conducting antimony doped tin oxide
nanoparticles. Chem. Mater. 27, 1090–1099. (doi:10.1021/cm504409k)
2. McManus JS, Cunningham PD, Regan LB, Smith A, McGrath DW, Dunne PW. 2014 Highly
soluble ligand stabilized tin oxide nanocrystals: gel formation and thin film production.
Crystal. Growth Des. 14, 4819–4826. (doi:10.1021/cg5009957)
3. Ito D, Yokoyama S, Zaikova T, Masuko K, Hutchison JE. 2014 Synthesis of ligand-stabilized
metal oxide nanocrystals and epitaxial core/shell nanocrystals via a lower-temperature
esterification process. ACS Nano 8, 64–75. (doi:10.1021/nn401888h)
4. Bob B, Song T-B, Chen C-C, Xu Z, Yang Y. 2013 Nanoscale dispersions of gelled SnO2 :
material properties and device applications. Chem. Mater. 25, 4725–4730. (doi:10.1021/cm
402462m)
5. Chen L, Fleming P, Morris V, Holmes JD, Morris MA. 2010 Size-related lattice parameter
changes and surface defects in ceria nanocrystals. J. Phys. Chem. C 114, 12 909–12 919.
(doi:10.1021/jp1031465)
6. Migani A, Vayssilov GN, Bromley ST, Illas F, Neyman KM. 2010 Dramatic reduction of
the oxygen vacancy formation energy in ceria particles: a possible key to their remarkable
reactivity at the nanoscale. J. Mater. Chem. 20, 10 535–10 546. (doi:10.1039/c0jm01908a)
7. Talapin DV, Lee J-S, Kovalenko MV, Shevchenko EV. 2009 Prospects of colloidal nanocrystals
for electronic and optoelectronic applications. Chem. Rev. 110, 389–458. (doi:10.1021/
cr900137k)
8. Michalet X, Pinaud FF, Bentolila LA, Tsay JM, Doose S, Li JJ, Sundaresen G, Gambhir SS, Weiss
S. 2005 Quantum dots for live cells, in vivo imaging, and diagnostics. Science 307, 538–544.
(doi:10.1126/science.1104274)
9. Burda C, Chen X, Narayanan R, El-Sayed MA. 2005 Chemistry and properties of nanocrystals
of different shapes. Chem. Rev. 105, 1025–1102. (doi:10.1021/cr030063a)
10. O’Regan B, Gratzel M. 1991 A low-cost, high-efficiency solar cell based on dye-sensitized
colloidal TiO2 films. Nature 353, 737–739. (doi:10.1038/353737a0)
 11. Wu P, Yan X-P. 2013 Doped quantum dots for chemo/biosensing and bioimaging. Chem. Soc.
18
Rev. 42, 5489–5521. (doi:10.1039/c3cs60017c)
12. Tan Z et al. 2007 Bright and color-saturated emission from blue light-emitting diodes

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
based on solution-processed colloidal nanocrystal quantum dots. Nano Lett. 7, 3803–3807.
(doi:10.1021/nl072370s)
13. Narayanan R, El-Sayed MA. 2005 Catalysis with transition metal nanoparticles in colloidal
solution: nanoparticle shape dependence and stability. J. Phys. Chem. B 109, 12 663–12 676.
(doi:10.1021/jp051066p)
14. Jing L, Zhou W, Tian G, Fu H. 2013 Surface tuning for oxide-based nanomaterials as efficient
photocatalysts. Chem. Soc. Rev. 42, 9509–9549. (doi:10.1039/c3cs60176e)
15. Chen G, Seo J, Yang C, Prasad PN. 2013 Nanochemistry and nanomaterials for photovoltaics.
Chem. Soc. Rev. 42, 8304–8338. (doi:10.1039/c3cs60054h)
16. Lee KT, Cho J. 2011 Roles of nanosize in lithium reactive nanomaterials for lithium ion
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

batteries. Nano Today 6, 28–41. (doi:10.1016/j.nantod.2010.11.002)

17. Zhang Q, Wang C-F, Ling L-T, Chen S. 2014 Fluorescent nanomaterial-derived white light-
emitting diodes: what’s going on. J. Mater. Chem. C 2, 4358–4373. (doi:10.1039/c4tc00048j)
18. Miller DR, Akbar SA, Morris PA. 2014 Nanoscale metal oxide-based heterojunctions for gas
sensing: a review. Sens. Actuators B 204, 250–272. (doi:10.1016/j.snb.2014.07.074)
19. Rao CNR. 1993 Chemical synthesis of solid inorganic materials. Mater. Sci. Eng. B Solid State
Mater. Adv. Technol. 18, 1–21. (doi:10.1016/0921-5107(93)90109-Z)
20. Choy KL. 2003 Chemical vapour deposition of coatings. Prog. Mater. Sci. 48, 57–170.
(doi:10.1016/S0079-6425(01)00009-3)
21. Helmersson U, Lattemann M, Bohlmark J, Ehiasarian AP, Gudmundsson JT. 2006 Ionized
physical vapor deposition (IPVD): a review of technology and applications. Thin Solid Films
513, 1–24. (doi:10.1016/j.tsf.2006.03.033)
22. Tjong SC, Chen H. 2004 Nanocrystalline materials and coatings. Mater. Sci. Eng. R Rep. 45,
1–88. (doi:10.1016/j.mser.2004.07.001)
23. Cushing BL, Kolesnichenko VL, O’Connor CJ. 2004 Recent advances in the liquid-
phase syntheses of inorganic nanoparticles. Chem. Rev. 104, 3893–3946. (doi:10.1021/cr03
0027b)
24. Eastoe J, Hollamby MJ, Hudson L. 2006 Recent advances in nanoparticle synthesis
with reversed micelles. Adv. Colloid Interface Sci. 128–130, 5–15. (doi:10.1016/j.cis.2006.
11.009)
25. Byrappa K, Yoshimura M. 2008 Handbook of hydrothermal technology. Amsterdam, The
Netherlands: Elsevier Science.
26. Modeshia DR, Walton RI. 2010 Solvothermal synthesis of perovskites and pyrochlores:
crystallisation of functional oxides under mild conditions. Chem. Soc. Rev. 39, 4303–4325.
(doi:10.1039/b904702f)
27. LaMer VK, Dinegar RH. 1950 Theory, production and mechanism of formation of
monodispersed hydrosols. J. Am. Chem. Soc. 72, 4847–4854. (doi:10.1021/ja01167a001)
28. Sugimoto T. 1987 Preparation of monodispersed colloidal particles. Adv. Colloid Interface Sci.
28, 65–108. (doi:10.1016/0001-8686(87)80009-X)
29. de Mello Donegá C, Liljeroth P, Vanmaekelbergh D. 2005 Physicochemical evaluation of the
hot-injection method, a synthesis route for monodisperse nanocrystals. Small 1, 1152–1162.
(doi:10.1002/smll.200500239)
30. Kwon SG, Hyeon T. 2011 Formation mechanisms of uniform nanocrystals via hot-injection
and heat-up methods. Small 7, 2685–2702. (doi:10.1002/smll.201002022)
31. Weingärtner H, Franck EU. 2005 Supercritical water as a solvent. Angew. Chem. Int. Ed. 44,
2672–2692. (doi:10.1002/anie.200462468)
32. Akiya N, Savage PE. 2002 Roles of water for chemical reactions in high-temperature water.
Chem. Rev. 102, 2725–2750. (doi:10.1021/cr000668w)
33. Adschiri T, Kanazawa K, Arai K. 1992 Rapid and continuous hydrothermal crystallization of
metal oxide particles in supercritical water. J. Am. Ceram. Soc. 75, 1019–1022. (doi:10.1111/
j.1151-2916.1992.tb04179.x)
34. Blood PJ, Denyer JP, Azzopardi BJ, Poliakoff M, Lester E. 2004 A versatile flow visualisation
technique for quantifying mixing in a binary system: application to continuous supercritical
water hydrothermal synthesis (SWHS). Chem. Eng. Sci. 59, 2853–2861. (doi:10.1016/j.ces.2004.
04.021)
 35. Lester E, Blood P, Denyer J, Giddings D, Azzopardi B, Poliakoff M. 2006 Reaction engineering:
19
the supercritical water hydrothermal synthesis of nano-particles. J. Supercrit Fluids 37,
209–214. (doi:10.1016/j.supflu.2005.08.011)

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
36. Hayashi H, Hakuta Y. 2010 Hydrothermal synthesis of metal oxide nanoparticles in
supercritical water. Materials 3, 3794–3817. (doi:10.3390/ma3073794)
37. Adschiri T, Lee Y-W, Goto M, Takami S. 2011 Green materials synthesis with supercritical
water. Green Chem. 13, 1380–1390. (doi:10.1039/c1gc15158d)
38. Hakuta Y, Onai S, Terayama H, Adschiri T, Arai K. 1998 Production of ultra-fine ceria particles
by hydrothermal synthesis under supercritical conditions. J. Mater. Sci. Lett. 17, 1211–1213.
(doi:10.1023/A:1006597828280)
39. Lester E, Aksomaityte G, Li J, Gomez S, Gonzalez-Gonzalez J, Poliakoff M. 2012 Controlled
continuous hydrothermal synthesis of cobalt oxide (Co3 O4 ) nanoparticles. Prog. Cryst. Growth
Charact. Mater. 58, 3–13. (doi:10.1016/j.pcrysgrow.2011.10.008)
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

40. Hakuta . Y, Seino K, Ura H, Adschiri T, Takizawa H, Arai K. 1999 Production of phosphor
(YAG .. Tb) fine particles by hydrothermal synthesis in supercritical water. J. Mater. Chem. 9,
2671–2674. (doi:10.1039/a903074c)
41. Cabanas A, Li J, Blood P, Chudoba T, Lojkowski W, Poliakoff M, Lester E. 2007 Synthesis of
nanoparticulate yttrium aluminum garnet in supercritical water–ethanol mixtures. J. Supercrit.
Fluids 40, 284–292. (doi:10.1016/j.supflu.2006.06.006)
42. Kaseda K, Takesue M, Aida TM, Watanabe M, Hayashi H, Smith Jr RL. 2011 Restructuring
mechanism of NbO6 octahedrons in the crystallization of KNbO3 in supercritical water.
J. Supercrit. Fluids 58, 279–285. (doi:10.1016/j.supflu.2011.06.009)
43. Helen R, Cyril A, Anne L-S, Catherine E, Mario M, François C. 2005 Single-step synthesis of
well-crystallized and pure barium titanate nanoparticles in supercritical fluids. Nanotechnology
16, 1137. (doi:10.1088/0957-4484/16/8/026)
44. Philippot G, Elissalde C, Maglione M, Aymonier C. 2014 Supercritical fluid technology:
a reliable process for high quality BaTiO3 based nanomaterials. Adv. Powder Technol. 25,
1415–1429. (doi:10.1016/j.apt.2014.02.016)
45. Dunne PW, Starkey CL, Gimeno-Fabra M, Lester EH. 2014 The rapid size- and shape-
controlled continuous hydrothermal synthesis of metal sulphide nanomaterials. Nanoscale 6,
2406–2418. (doi:10.1039/c3nr05749f)
46. Dunne PW, Munn AS, Starkey CL, Lester EH. 2015 The sequential continuous-flow
hydrothermal synthesis of molybdenum disulphide. Chem. Commun. 51, 4048–4050.
(doi:10.1039/C4CC10158H)
47. Aimable A, Aymes D, Bernard F, Le Cras F. 2009 Characteristics of LiFePO4 obtained through
a one step continuous hydrothermal synthesis process working in supercritical water. Solid
State Ionics 180, 861–866. (doi:10.1016/j.ssi.2009.02.019)
48. Hong S-A, Kim SJ, Chung KY, Lee Y-W, Kim J, Sang B-I. 2013 Continuous synthesis of lithium
iron phosphate nanoparticles in supercritical water: effect of process parameters. Chem. Eng.
J. 229, 313–323. (doi:10.1016/j.cej.2013.05.094)
49. Hong S-A, Kim SJ, Kim J, Lee BG, Chung KY, Lee Y-W. 2012 Carbon coating on lithium iron
phosphate (LiFePO4 ): comparison between continuous supercritical hydrothermal method
and solid-state method. Chem. Eng. J. 198–199, 318–326. (doi:10.1016/j.cej.2012.05.058)
50. Lester E, Tang SVY, Khlobystov A, Rose VL, Buttery L, Roberts CJ. 2013 Producing nanotubes
of biocompatible hydroxyapatite by continuous hydrothermal synthesis. CrystEngComm 15,
3256–3260. (doi:10.1039/c3ce26798a)
51. Chaudhry AA, Haque S, Kellici S, Boldrin P, Rehman I, Khalid FA, Darr JA. 2006 Instant
nano-hydroxyapatite: a continuous and rapid hydrothermal synthesis. Chem. Commun. 4,
2286–2288. (doi:10.1039/b518102j)
52. Chaudhry AA, Goodall J, Vickers M, Cockcroft JK, Rehman I, Knowles JC, Darr JA. 2008
Synthesis and characterisation of magnesium substituted calcium phosphate bioceramic
nanoparticles made via continuous hydrothermal flow synthesis. J. Mater. Chem. 18,
5900–5908. (doi:10.1039/b807920j)
53. Aksomaityte G, Poliakoff M, Lester E. 2013 The production and formulation of
silver nanoparticles using continuous hydrothermal synthesis. Chem. Eng. Sci. 85, 2–10.
(doi:10.1016/j.ces.2012.05.035)
54. Choi H, Veriansyah B, Kim J, Kim J-D, Kang JW. 2010 Continuous synthesis of metal
nanoparticles in supercritical methanol. J. Supercrit. Fluids 52, 285–291. (doi:10.1016/j.supflu.
2010.01.015)
 55. Kim J, Kim D, Veriansyah B, Won Kang J, Kim J-D. 2009 Metal nanoparticle synthesis using
20
supercritical alcohol. Mater. Lett. 63, 1880–1882. (doi:10.1016/j.matlet.2009.05.066)
56. Kubota S, Morioka T, Takesue M, Hayashi H, Watanabe M, Smith Jr RL. 2014 Continuous

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
supercritical hydrothermal synthesis of dispersible zero-valent copper nanoparticles
for ink applications in printed electronics. J. Supercrit Fluids 86, 33–40. (doi:10.1016/
j.supflu.2013.11.013)
57. Gimeno-Fabra M, Munn AS, Stevens LA, Drage TC, Grant DM, Kashtiban RJ, Sloan J, Lester
E, Walton RI. 2012 Instant MOFs: continuous synthesis of metal–organic frameworks by rapid
solvent mixing. Chem. Commun. 48, 10 642–10 644. (doi:10.1039/c2cc34493a)
58. Kanamura K, Goto A, Young Ho R, Umegaki T, Toyoshima K, Okada KI, Hakuta Y,
Adschiri T, Arai K. 2000 Preparation and electrochemical characterization of LiCoO2
particles prepared by supercritical water synthesis. Electrochem. Solid-State Lett. 3, 256–258.
(doi:10.1149/1.1391117)
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

59. Hakuta Y, Adschiri T, Hirakoso H, Arai K. 1999 Chemical equilibria and particle morphology
of boehmite (AlOOH) in sub- and supercritical water. Fluid Phase Equilib. 158, 733–742.
(doi:10.1016/S0378-3812(99)00118-1)
60. Shaw WHR, Walker DG. 1956 The decomposition of thiourea in water solutions. J. Am. Chem.
Soc. 78, 5769–5772. (doi:10.1021/ja01603a014)
61. Erickson BE, Helz GR. 2000 Molybdenum(VI) speciation in sulfidic waters: stability and
lability of thiomolybdates. Geochim. Cosmochim. Acta 64, 1149–1158. (doi:10.1016/S0016-
7037(99)00423-8)
62. Hibble SJ, Wood GB. 2003 Modeling the structure of amorphous MoS3 : a neutron diffraction
and reverse Monte Carlo study. J. Am. Chem. Soc. 126, 959–965. (doi:10.1021/ja037666o)
63. Whittingham MS. 2012 History, evolution, and future status of energy storage. Proc. IEEE 100,
1518–1534. (doi:10.1109/JPROC.2012.2190170)
64. Padhi AK, Nanjundaswamy KS, Masquelier C, Okada S, Goodenough JB. 1997 Effect
of structure on the Fe3+ /Fe2+ redox couple in iron phosphates. J. Electrochem. Soc. 144,
1609–1613. (doi:10.1149/1.1837649)
65. Zhang W-J. 2011 Structure and performance of LiFePO4 cathode materials: a review. J. Power
Sources 196, 2962–2970. (doi:10.1016/j.jpowsour.2010.11.113)
66. Dreaden EC, Alkilany AM, Huang X, Murphy CJ, El-Sayed MA. 2012 The golden age: gold
nanoparticles for biomedicine. Chem. Soc. Rev. 41, 2740–2779. (doi:10.1039/C1CS15237H)
67. Li Y, Liu Q, Shen W. 2011 Morphology-dependent nanocatalysis: metal particles. Dalton Trans.
40, 5811–5826. (doi:10.1039/c0dt01404d)
68. Lee KJ, Jun BH, Kim TH, Joung J. 2006 Direct synthesis and inkjetting of silver
nanocrystals toward printed electronics. Nanotechnology 17, 2424. (doi:10.1088/0957-4484/17/
9/060)
69. Park BK, Kim D, Jeong S, Moon J, Kim JS. 2007 Direct writing of copper conductive patterns
by ink-jet printing. Thin Solid Films 515, 7706–7711. (doi:10.1016/j.tsf.2006.11.142)
70. Yu J, Savage PE. 1998 Decomposition of formic acid under hydrothermal conditions. Ind. Eng.
Chem. Res. 37, 2–10. (doi:10.1021/ie970182e)
71. Long JR, Yaghi OM. 2009 The pervasive chemistry of metal–organic frameworks. Chem. Soc.
Rev. 38, 1213–1214. (doi:10.1039/b903811f)
72. Jhung S-H, Lee J-H, Chang J-S. 2005 Microwave synthesis of a nanoporous hybrid
material, chromium trimesate. Bull. Korean Chem. Soc. 26, 880–881. (doi:10.5012/bkcs.2005.26.
6.880)
73. Klinowski J, Almeida Paz FA, Silva P, Rocha J. 2011 Microwave-assisted synthesis of metal–
organic frameworks. Dalton Trans. 40, 321–330. (doi:10.1039/C0DT00708K)
74. Ni Z, Masel RI. 2006 Rapid production of metal–organic frameworks via microwave-assisted
solvothermal synthesis. J. Am. Chem. Soc. 128, 12 394–12 395. (doi:10.1021/ja0635231)
75. Seo Y-K, Hundal G, Jang IT, Hwang YK, Jun C-H, Chang J-S. 2009 Microwave synthesis
of hybrid inorganic–organic materials including porous Cu3 (BTC)2 from Cu(II)–trimesate
mixture. Micropor. Mesopor. Mater. 119, 331–337. (doi:10.1016/j.micromeso.2008.10.035)
76. Son W-J, Kim J, Kim J, Ahn W-S. 2008 Sonochemical synthesis of MOF-5. Chem. Commun. 2008,
6336–6338. (doi:10.1039/b814740j)
77. Klimakow M, Klobes P, Thünemann AF, Rademann K, Emmerling F. 2010 Mechanochemical
synthesis of metal−organic frameworks: a fast and facile approach toward quantitative yields
and high specific surface areas. Chem. Mater. 22, 5216–5221. (doi:10.1021/cm1012119)
 78. Schoenecker PM, Belancik GA, Grabicka BE, Walton KS. 2013 Kinetics study and
21
crystallization process design for scale-up of UiO-66-NH2 synthesis. AlChE J 59, 1255–1262.
(doi:10.1002/aic.13901)

rsta.royalsocietypublishing.org Phil. Trans. R. Soc. A 373: 20150015

.........................................................
79. Garcia Marquez A, Horcajada P, Grosso D, Ferey G, Serre C, Sanchez C, Boissiere C. 2013
Green scalable aerosol synthesis of porous metal–organic frameworks. Chem. Commun. 49,
3848–3850. (doi:10.1039/c3cc39191d)
80. Carné-Sánchez A, Imaz I, Cano-Sarabia M, Maspoch D. 2013 A spray-drying strategy
for synthesis of nanoscale metal–organic frameworks and their assembly into hollow
superstructures. Nat. Chem. 5, 203–211. (doi:10.1038/nchem.1569)
81. Faustini M, Kim J, Jeong G-Y, Kim JY, Moon HR, Ahn W-S, Kim D-P. 2013 Microfluidic
approach toward continuous and ultrafast synthesis of metal–organic framework crystals
and hetero structures in confined microdroplets. J. Am. Chem. Soc. 135, 14 619–14 626.
(doi:10.1021/ja4039642)
Downloaded from https://royalsocietypublishing.org/ on 28 July 2021

82. Kim K-J, Li YJ, Kreider PB, Chang C-H, Wannenmacher N, Thallapally PK, Ahn H-G. 2013
High-rate synthesis of Cu–BTC metal–organic frameworks. Chem. Commun. 49, 11 518–11 520.
(doi:10.1039/c3cc46049e)
83. Paseta L, Seoane B, Julve D, Sebastián V, Téllez C, Coronas J. 2013 Accelerating the controlled
synthesis of metal–organic frameworks by a microfluidic approach: a nanoliter continuous
reactor. ACS Appl. Mater. Interfaces 5, 9405–9410. (doi:10.1021/am4029872)
84. Rubio-Martinez M, Batten MP, Polyzos A, Carey K-C, Mardel JI, Lim K-S, Hill MR. 2014
Versatile, high quality and scalable continuous flow production of metal–organic frameworks.
Sci. Rep. 4, 5443. (doi:10.1038/srep05443)
85. D’Arras L, Sassoye C, Rozes L, Sanchez C, Marrot J, Marre S, Aymonier C. 2014 Fast and
continuous processing of a new sub-micronic lanthanide-based metal–organic framework.
New J. Chem. 38, 1477–1483. (doi:10.1039/c3nj01371e)
86. Bayliss PA, Ibarra IA, Perez E, Yang S, Tang CC, Poliakoff M, Schröder M. 2014
Synthesis of metal–organic frameworks by continuous flow. Green Chem. 16, 3796–3802.
(doi:10.1039/C4GC00313F)
87. Waitschat S, Wharmby MT, Stock N. 2015 Flow-synthesis of carboxylate and phosphonate
based metal–organic frameworks under non-solvothermal reaction conditions. Dalton Trans.
44, 11 235–11 240. (doi:10.1039/C5DT01100K)
88. Biemmi E, Christian S, Stock N, Bein T. 2009 High-throughput screening of synthesis
parameters in the formation of the metal–organic frameworks MOF-5 and HKUST-1. Micropor.
Mesopor. Mater. 117, 111–117. (doi:10.1016/j.micromeso.2008.06.040)
89. Bøjesen ED, Jensen KMø, Tyrsted C, Lock N, Christensen M, Iversen BB. 2014 In situ powder
diffraction study of the hydrothermal synthesis of ZnO nanoparticles. Crystal. Growth Des. 14,
2803–2810. (doi:10.1021/cg5000606)
90. Breraholm M, Jensen H, Iversen SB, Iversen BB. 2008 Reactor design for in situ X-ray scattering
studies of nanoparticle formation in supercritical water syntheses. J. Supercrit. Fluids 44,
385–390. (doi:10.1016/j.supflu.2007.09.029)
91. Middelkoop V, Boldrin P, Peel M, Buslaps T, Barnes P, Darr JA, Jacques SDM. 2009 Imaging
the inside of a continuous nanoceramic synthesizer under supercritical water conditions using
high-energy synchrotron X-radiation. Chem. Mater. 21, 2430–2435. (doi:10.1021/cm900118z)