INTRODUCTION

Global warming, energy crisis, and pollution are serious concerns affecting both the human health
and environment. The environmental pollution includes a wide range of hazardous chemicals
which are harmful even at extremely low concentrations.

The common treatment of these pollutants involves the use of pyrolytic methods which
consume large amounts of fossil fuels producing elevated levels of CO 2 and thus contributing
again to climate change and energy crisis.
Another severe issue [1–3] is water pollution from emerging contaminants (ECs) such as endocrine
disrupting chemicals, pharmaceuticals, and personal care products even at trace levels. Some
adverse potential effects caused by ECs are water toxicity, resistance development in pathogenic
bacteria, genotoxicity, and endocrine disruption [4–6].
Wastewater treatment plants are not designed to remove low concentrations of synthetic
pollutants such as pharmaceuticals and hence alternatives such us the Advanced Oxidation
Technology (AOT) have been used to solve this environmental problem [7, 8].
Heterogeneous photocatalytic processes constitute one of the most important AOTs and are based
on the oxidation of polluting compounds which can be found in air or water by means of a
reaction occurring on a semiconductor catalytic surface activated by light with a specific
wavelength.
TiO2 is the most investigated semiconductor catalyst particularly because of its great potential in
the treatment of environmental pollution [9] and it is chemically stable, nontoxic, and inexpensive
[10–12]. However, TiO2 has an important disadvantage as it is usually commercially available as
powder and, therefore, a posttreatment separation stage is needed [13, 14] and hence it would be
highly desirable to have the possibility of supporting the TiO2 active phase on other structured
inorganic materials.
In this sense, a new exciting approach for the production of structured materials on demand may
be the use of a 3D printing (3DP) system. Until the past few years, this type of technology has
been restricted to medium- and big-sized
 FIgURe 1: Indirect Inject Print (IIP) is a layer-by-layer process of depositing liquid binder
onto thin layers of powder to create a 3D object [17].

companies devoted to the fabrication of prototypes. Nowa- days, they are increasingly available
for small businesses (research [15]) and even for direct manufacturing. 3D micro fabrication
remained a challenge until the introduction of free-forming fabrication (FFF) technology. Free-
forming fab- rication is any fabrication technology that fabricates 3D com- plex structures by
assembling small elements together and usually starts with a computer aided design (CAD)
model. FFF includes, but is not limited to, rapid prototyping, 3D printing, and direct writing for
macro scale fabrication [16].

A digital model of the object is created in a computer. Using adequate software, the user may
control several rel- evant parameters such as the number of layers which are piled up to generate
the full item depending on the resolution required, thickness of the layers, and porosity. In the
second step, each digital layer is printed in an appropriate substrate. Different 3D printing
techniques can be selected depending on the material required and the way to fuse layers together.
One of the 3D printing options is the Indirect Inject Print, where a powder is spread from a well
(Figure 1), levelling it to produce a thin layer. Then the printer heads will dispense a thin layer of
a binder in the required pattern of the cross section. When the layer is finished, the “build tray”
will be lowered by a fraction of a millimetre, typically between 10 and 100 𝜇 and then the process is
repeated again for the next cross section. Finally, after printing the whole 3D structure, the loose
powder is blown away with compressed air, revealing the full structure [16].

This type of technology has the potential to allow the fabrication of monolithic porous structures
that can be covered with a layer of catalytically active material such as TiO2 [18]. Calcium
sulphate hemihydrate (CaSO4 ⋅1/2H2 O) was one of the first materials to be used for IIP. It can be
 wetted using a commercially formulated binder (98% content water) and then forms a gypsum
paste (CaSO4 ⋅2H2 O) by activating self-hydration [19] which can be used to produce 3D porous
supporting structures.

MATERIALS AND METHODS

The activity of the TiO2 photocatalyst supported over a 3D printed ceramic structure was studied
evaluating the pho- todegradation of methylene blue (Sigma-Aldrich) as a model molecule for
wastewater treatments. The photocatalytically active material, TiO2 (Degussa P25, Evonik
Industries), was examined by X-ray diffraction (XRD) in order to obtain the percentage of
anatase and rutile phase. Nitrogen adsorption- desorption porosimetry and mercury porosimetry
tech- niques were used to study its specific surface area and textural properties [20]. TiO2
(Degussa P25) was mainly composed by the photocatalytically active anatase phase, 81% (wt).
The textural parameters of TiO2 powder were 51.1 m2/g of BET specific area, a total pore area of 63.7
m2/g with 92.5% of porosity, and a density of 2.58 g/ml.

3D Printing Structures.

The materials used in the 3D printing process were calcium sulphate hemihydrate (VisiJet PXL

Core, from S.A.T.E´ LITE) and a water-based binder (VisiJet Clear, from S.A.T.E´ LITE). These

materials were used in a 3D printer (ProJet® 360, from S.A.T.E´ LITE), with a 300

× 450 dpi resolution and layer thickness of 0.1 mm.

First, the structure is digitally designed using free soft-

ware Tinkercad [22] and the corresponding digital model is saved as a stl file and then sent to
the 3D printer. During the three printing processes (Figures 2(a)–2(d)), a roller spread a thin
powder layer from the feed area to the build area and then the print head deposits binder
droplets selectively within the build area. When the first layer is printed, the roller returns to the
feed area, spreading another powder layer to the build area. This procedure is repeated until the
fabrication of the whole 3D structure is completed [21].
Following this procedure, several microtubes of ceramic supporting structures were 3D
printed (Figures 3(a) and 3(b)), and several small rectangular and square pieces of 2 × 2 cm
were also 3D printed as testing samples in mechanical stability versus temperature studies
 (Figures 3(c) and 3(d)).
150∘C was the selected prefiring temperature to give enough mechanical stability for
withstanding the conditions of the impregnation process.

Improving the Mechanical Stability of the 3D Printed Structures.

Two routes were followed to impregnate the 3D structures with another inorganic material
enhancing the mechanical stability up to 1200∘C. The 3D printed samples were previously fired at
150∘C for2 h to improve the mechan- ical stability in aqueous solution needed for the
impregnation processes.

Route 1. Impregnation was done with an aqueous dis- persion of alumina (Al2O3, Almatis GmbH)
using Dolapix CE64 (Zschimmer & Schwarz) as dispersant. Several ratios of the components and
impregnation times (Table 1) were considered to optimize the process. The impregnation times
were also optimized ranging from 10 s up to 30 min. The so- impregnated samples were left to dry
for 45 min under 70∘C and then were finally fired at 1200∘C for4 h with heating and cooling ramp
rates of 5∘C/min (Figure 4(a)).

Route 2. Impregnation was done with an aqueous solution of Ludox® AS-30 colloidal silica (Sigma-
Aldrich). As in the previous method, the printed 3D structures were initially fired at 150∘C for 2 h.
After that, the specimens were impreg- nated in Ludox (1.1 g) and then fired at 600 ∘C for 2 h, with
heating and cooling ramp rates of 5∘C/min (Figure 4(b)); this step is performed twice.

Photocatalytic Studies. The methylene blue (MB) photo- degradation process with the activated
3D printed structures (A3DS) was carried out in a fixed-bed photoreactor system
 Feed area
Fe ea
Build area Build area
ed ar

Fe
Feed Build
ar Build ed
area

FIgURe 2: 3DP process scheme. (a) The roller spreads a thin layer of powder from the feed area
to the build area. (b) The print head injects binder droplets on the powder bed. (c) After
printing a layer, the roller returns to the feed area. (d) Powder in the feed area is raised, while
that in the build area is lowered. The roller then spreads another layer of powder [21].

2 mm

(c) (d)

1
cm
FIgURe 3: ((a) and (b)) 3D printed supporting structures in CaSO4. (c) 3D printed square testing
pieces in the green state and (d) after firing at 250∘C for3 h.

(1) Printing 3D (3) Impregantion (4) Dried under

(2) Prefiring at (5) Firing at
ceramic
supports with A Ｆ2 ／3 air, at 70 C, for
∘
150∘C, 2 h 1200∘C, 4 h

(a)

(1) Printing 3D (4) Dried under (5) Firing at

(2) Prefiring at (3) Impregantion
ceramic air for 24 h, 600∘C, 2 h, and
supports with Ludox then repeat (1)–
150 C, 2 h
∘
(4)
room T

(b)

FIgURe 4: Schemes for improving the mechanical stability of 3D printed supporting structures
using two alternative impregnation routes: (a) route 1 with Al2O3 and (b) route 2 with Ludox
dispersions.

(Figure 5(a)). A3DS were placed in the photocatalytic reactor (18 cm length and 0.6 cm internal
diameter) (Figure 5(b)), and a solar radiation sodium vapor lamp (Philips, model 400-WG/92/2),
placed at a distance of 50 cm from the fixed- bed reactor, was used as light source (total
radiation flux measured on the fixed-bed surface was 160 mW/cm 2). MB solution in water (20
ppm) was introduced in a 250 ml wastewater photoreactor tank (Figure 5(c)), keeping the tem-
perature constant at 25∘C. The MB solution was recirculated
along the system using a peristaltic pump (Watson-Marlow, model 302S). Several samples were
taken during several hours of irradiation time and they were analyzed by UV-Vis spec-
trophotometry in order to follow the evolution of MB concen- tration into the reactor.
Moreover, MB photolysis (with light source and without photocatalyst) and MB adsorption
onto A3DS (with photocatalyst and without light source) experi- ments were developed in
order to evaluate its contribution to the global wastewater decontamination process.

Microstructural Characterization. The morphologies of the 3D structured samples in the

green and sintered states were examined using a stereomicroscope Leica Zoom 2000 (Leica
Microsystems, Inc.) and a scanning electron micro- scope (SEM) (model Jeol LTD, JSM-6300)
combined with energy dispersive spectroscopy (EDS). For SEM studies, all samples were
covered with a thin film of sputtered silver to avoid charging problems and to obtain better
image definition.
 CONCLUSION

3D printing technology has been successfully used to create 3D supporting structures of CaSO4
which can be activated with TiO2 showing photocatalytic activity Two approaches have been
proposed and optimized to enhance the mechanical stability of the 3D supports up to 1200∘C. In
one of them, Al2O3 is used as a binder of the whole 3D structure; in the other method, a
dispersion of nanoparticles of SiO2 (Ludox) acts as a binder.
In both cases, the A3DS show photocatalytic activity for the MB removal from wastewater,
reaching more than 50% of conversion in just 1 h of irradiation when using Ludox in the
activation process. This approach reaches a high photodegradation rate of about 90% after 5 h.
The A3DS offer clear advantages for the industrial treat- ment of wastewater given that the
photocatalytic materials remain confined to the packed bed avoiding the need of sep- arating the
catalyst from the decontaminated water effluent.