Solar Energy 273 (2024) 112496

Contents lists available at ScienceDirect

Solar Energy
journal homepage: www.elsevier.com/locate/solener

A comparative study of acrylic and epoxy-based adhesives for perovskite

solar cells encapsulation
Gabriela A. Soares a, Izabela S. Bicalho a, Sergio Castro-Hermosa a, b, Luiza de Q. Corrêa a,
Bárbara H.S. Miranda a, Adriano dos S. Marques a, Silvia L. Fernandes a, c, Thainá Cunha a,
Vinícius V. de Freitas a, Rodrigo de Q. Vilaça a, Luana Wouk a, d, *, Diego Bagnis a
a
Oninn, Printed Electronics Department, Rua Sete, 2000, Belo Horizonte, Minas Gerais 31035-536, Brazil
b
Universidad El Bosque, Facultad de Ingeniería, Grupo de investigación OSIRIS&BIOAXIS, Av. Cra 9 No. 131 A – 02, Bogotá, DC, Colombia
c
Universidade Estadual Paulista Julio de Mesquita Filho, Faculdade de Ciências, Campus BauruAv. Eng. Luiz Edmundo Carrijo Coube, 14-01, Bauru, São Paulo 17033-
360, Brazil
d
Institute of Physics, University of Brasilia, Federal District, 70910-900 Brasilia, Brazil

A R T I C L E I N F O A B S T R A C T

Keywords: The development of a reliable encapsulation is a key factor for protecting perovskite solar cells from extrinsic
Perovskite solar cell degradation and an important step to commercialization. The use of edge sealing materials has already been
Encapsulation proven effective for rigid substrates. However, for flexible devices, a different encapsulation strategy must be
Epoxy
developed. In this study, epoxy- and acrylic-based adhesives were tested over blade-coated p-i-n MAPI devices
Acrylic
and the encapsulation was evaluated with current–voltage measurements, calcium, and thermal stability tests.
Although causing discoloration around the cells, over the area without the evaporated top electrode, the epoxy
adhesive showed great performance after encapsulation and under thermal stress. This strategy was used to
compare the stability of PEDOT:PSS and NiO as HTL. After an initial drop of 40 % in performance, the device
with NiO was stable for over 5500 h under 85 ◦ C. These results show that this method can be used for evaluating
the stability of perovskite solar cell and that, with further development of a proper buffer layer, epoxy-based
adhesives are an effective strategy for flexible encapsulation.

1. Introduction is mainly related to external environmental factors such as moisture,

oxygen, and incident light. The perovskite material, i.e. CH3NH3PbI3,
Perovskite solar cells (PSC) have received extensive attention in the tends to decompose in the presence of water or oxygen and the degra­
last few years. The technology has developed fast, reaching laboratory dation is accelerated under illumination [6,7]. Fortunately, it has been
cell and module power conversion efficiencies (PCE) of 25.5 % and 17.9 proved that extrinsic factors can be effectively avoided by good encap­
% efficiency, respectively, in just one decade [1,2]. Despite the sulation strategies.
remarkable advances, there is still a long way to go before they can be Several encapsulation studies have been reported in the literature in
commercially viable. One of the biggest challenges is related to reli­ recent years, using a variety of materials and processes: hotmelt films,
ability, since PSCs are widely known to be unstable under the common including ethylene vinyl acetate (EVA) [8–10], polyurethane (PU) [8]
stressors an installation would face: light, heat, oxygen, moisture, and polyolefin (POE) [8,11]; polyisobutylene (PIB) and butyl rubber
electrical bias, among others [3–5]. Part of this instability is connected [11,12]; Surlyn ionomer [10,13]; poly(methyl methacrylate) (PMMA)
to intrinsic factors that depend mainly on the materials/composition [14,15]; Polydimethylsiloxane (PDMS) [16]; UV-curable adhesives
used and environmental conditions involved in the manufacturing [13,17–19]; Paraffin [20]; Al2O3, deposited by atomic layer deposition
process, that affect the crystallization of the perovskite and the density (ALD) [21–23]; glass frit and hermetical seal [24,25]; by adding desic­
of lattice defects, inducing degradation by ion migration and molecular cants [19], among others. Despite the significant number of results, it is
dissociation [3,6]. Meanwhile, the extrinsic component of the instability difficult to make comparisons between the reports, since different cell

* Corresponding author.
E-mail address: [email protected] (L. Wouk).

https://doi.org/10.1016/j.solener.2024.112496
Received 19 September 2023; Received in revised form 21 February 2024; Accepted 26 March 2024
Available online 11 April 2024
0038-092X/© 2024 International Solar Energy Society. Published by Elsevier Ltd. All rights reserved.
G.A. Soares et al. Solar Energy 273 (2024) 112496

Fig. 1. (a) Structure of the p-i-n cell and (b) picture of the final device.

Fig. 2. Steps of the small devices encapsulation: (a) the adhesive is dropped on top of the device using a syringe, then (b) a piece of glass is placed on top and
manually pressured to spread the adhesive over the whole surface. (c) the device is then exposed to a UV lamp for initiating the cure, which is completed after 24 h
under ambient temperature.

structures and perovskite formulations were used, as well as testing acid methyl ester (PCBM) was purchased from Nano-C and nickel
conditions. oxide (NiO) nanoparticles (2.5 wt% in ethanol) was purchase from
In general, the encapsulation of solar cells can be carried out Avantama. All reagents listed in this section were used without any
following two different strategies: by depositing a protective thin layer further purification. Two versions of commercial epoxy and acrylic
on top of the device or by sealing the edges. The second option is the liquid adhesives were used for the encapsulation. The four adhesives
most practical but is limited to rigid substrates. Flexible devices require tested were labeled as Acrylic A, Acrylic B, Epoxy A and Epoxy B. The
thin film encapsulation or flexible barrier and adhesives. In this case, the adhesives have high water vapor barrier and are UV/Vis-cured. Acrylic B
encapsulant must cover the whole area and, therefore, be chemically is a more flexible adhesive, when compared to Acrylic A, and epoxy B,
compatible with the top layers, mainly with the last deposited layer. unlike Epoxy A, can also be cured thermally, without UV–Vis exposure.
Additionally, adhesives must have thermal expansion coefficients More information about the adhesives can be found in the SI.
similar to the solar cell layers, be highly transparent, thermally and UV-
stable, with a very low oxygen and water vapor transmission rates
(WVTR) and processable at low temperatures [26–29]. In this study, two 2.2. Device fabrication
versions of liquid UV/Vis-curable acrylic- and epoxy-based adhesives
were tested for encapsulating p-i-n MAPI perovskite solar cells. Labo­ Samples were fabricated following process reported elsewhere [30].
ratory devices processed in air by a meniscus assisted technique were ITO coated glass substrates (10 Ohm/Sq) were cleaned with Liquinox (2
investigated by using PEDOT:PSS and NiO as hole transport layers % in H2O), deionized water, acetone, and IPA for 10 min each in an
comparing performance before and after encapsulation, and thermal ultrasound bath. The substrates were then dried with N2 and treated by
stability. ultraviolet ozone (NovaScan PSD 4) for 15 min. All layers were depos­
ited by blade coating technique except for the top counter electrode. All
2. Experimental section the blade coating steps, as well as the annealing, were carried out in air,
under controlled humidity below 40 % using a Erichsen 510 coater.
2.1. Materials First, a hole transporting material (PEDOT:PSS or NiO) was deposited on
top of cleaned ITO/Glass. PEDOT:PSS solution was diluted in IPA (1:3)
Indium tim oxide (ITO) coated glass substrate was purchased from and deposited at 10 mm/s and 60 ◦ C, using a gap height of 200 um.
Thin Film Devices (TFD). Poly(3,4-ethylenedioxythiophene) poly­ These films were annealed at 150 ◦ C for 10 min. The NiO nanoparticles
styrene sulfonate (PEDOT:PSS) (Clevious) was purchased from Heraeus. dispersion was diluted in a proportion of 1:5 in ethanol and deposited at
Methylammonium iodide (MAI) was purchased from Greatcell Solar 5 mm/s, 40 ◦ C and gap height of 575 μm. The films were annealed using
Materials. Lead (II) chloride (PbCl2), lead (II) acetate (Pb an infrared lamp (Heraeus medium-wave twin-tube CARBON infrared
(CH3CO2)2⋅3H2O), anhydrous dimethylformamide (DMF), anhydrous o- emitter CIR®) for 2 min followed by an UVO3 treatment for 2.5 min.
xylene, anhydrous ethanol (EtOH) and anhydrous isopropyl alcohol Perovskite ink was prepared by mixing Pb(CH3CO2)2⋅3H2O, PbCl2 and
(IPA) were purchased from Sigma Aldrich. Bathocuproine (BCP) was CH3NH3I with a molar ratio of 0.8:0.2:2.9 in DMF. The solution was
purchased from Tokyo Chemical Industry (TCI). Phenyl-C61-butyric stirred overnight at room temperature. Perovskite films were blade
coated at 43 ◦ C with a gap height of 100 μm at a coating speed of 10 mm/

2
G.A. Soares et al. Solar Energy 273 (2024) 112496

Fig. 3. (a) PCE, (b) Voc, (c) Jsc and (d) FF of perovskite solar cells before and after encapsulation with both versions of acrylic and epoxy adhesives. Solid lines
represent measurements in reverse scan, whilst dashed lines represent forward scan measurements.

Fig. 4. Pictures of encapsulated devices after 24 h.

s. The crystallization process was assisted by a nitrogen blower set at 20 the device at a rate of 1 Å/s using an Angstrom NexDep PVD system,
psi. PCBM films were deposited as electron transport layer (ETL) from a resulting in 8 cells of 0.5 cm2 active area. Cells architecture and a
20 mg/mL solution in o-xylene at 10 mm/s, temperature of 50 ◦ C and photograph of the final device are shown in Fig. 1.
gap height of 125 μm. BCP film was deposited from a 0.5 mg/mL solu­
tion in IPA at 5 mm/s, temperature of 40 ◦ C and gap height of 100 μm.
Finally, 100 nm of Ag electrode was thermally evaporated at the top of

3
G.A. Soares et al. Solar Energy 273 (2024) 112496

Fig. 5. Images of the edge of a cell (a) non-encapsulated and (b) encapsulated with epoxy, after total discoloration. In (a), all areas are covered with the perovskite
film; in (b), the perovskite layer around the cell is missing.

forward direction, with 100 points between –0.1 and 1.2 V and a settling
time of 25 ms.. Absorption spectra were measured with a UV Vis Spec­
trophotometer Shimadzu 2600 and the external quantum efficiency
(EQE) measurements were made with a PTS-2-QE Quantum Efficiency/
IPCE System from Sciencetech (See Figure S1 in Support Information).
An optical microscope (Leica, model DFC295) was used to assess visual
changes. The XRD spectra were obtained with Cu-Kalfa (λ = 1.54060 Å)
radiation source from Shimadzu XRD 7000. The source was operated at
40 kV and 30 mA. The incident angle was fixed at 3◦ . Photoluminescence
(PL) and Raman spectroscopy were performed with a confocal Witec
Alpha 300 system with a 15 mW and 438 nm excitation source.

2.5. Calcium test

A thin layer of >99 % calcium (approx. 200 nm) was deposited on

glass substrates of 50 mm x 50 mm by thermal evaporation under vac­
uum (10− 6 mbar) at a rate of 2 nm/s and encapsulated, following the
same procedures described for the PV devices. Due to the sensitivity of
calcium to water and oxygen, the encapsulation process in this case was
performed inside a nitrogen glove box.
Fig. 6. Evolution of absorbance spectra of ITO/PEDOT:PSS/Perovskite/PCBM/
BCP stack encapsulated with Epoxy A over time. Measurements made in the
central area between cells. 3. Results and discussion

2.3. Encapsulation 3.1. Encapsulation

The encapsulation process was held in ambient air conditions, using Preventing the damage of external factors as humidity, oxygen and
glass sheets of the same size of the substrates (50 x 50 mm). For con­ water is the key point of encapsulation. Most encapsulation processes
tacting the cells, tin-coated copper tapes from Adhesives Research® use UV curing, high temperature or pressure, which are harmful to the
were placed as bus bar on the positive and negative contacts of each devices resulting in efficiency losses after encapsulation. Here, samples
device. Next, the encapsulating adhesives were deposited (approx. 2 mL) were prepared without the application of high temperatures and pres­
on the center of the device using a syringe. Glass was placed on top and sure, with minimal UV irradiation. The electrical parameters of the de­
manually pressed to spread the adhesive through the whole area. vices before and after the encapsulation process are shown in Fig. 3. A
Average thickness of adhesives reached through this method are in the representative EQE spectrum of the cells before encapsulation is dis­
range of 120 μm for the epoxies and 75 μm for the acrylics. This dif­ played in SI Figure S1. The IV curves, subsequent to achieving a
ference is due to the lower viscosity of the acrylic versions. Devices were balanced load distribution are depicted in SI Figure S2. The J-V (cur­
then placed under a 395 nm UV lamp for 20 s in order to initialize the rent–voltage) characteristic shape of perovskite solar cells is often
curing process, which was completed after 24 h at room temperature. associated with recombination within the interlayer and PVK, possibly
The same process was used for all adhesives tested and is the one rec­ linked to the presence of oxygen vacancies, or it can be influenced by
ommended by the supplier for the curing to succeed. Schematics of the suboptimal PVK formation, which can introduce deep traps or free ions
encapsulation process are shown in Fig. 2. [31].
No significant drop in performance was observed with any of the
adhesives tested, suggesting that the fast UV cure was suitable for the
2.4. Characterization MAPI perovskite solar cells encapsulation. Visually, however, a major
difference could be observed: both devices with the epoxy adhesives
Cells were measure in a Class AAA solar simulator from Wacom, presented discoloration around the active area of the cells, which was
model WXS-156S-10, under standard conditions (filter AM 1.5G with an not observed with the acrylic ones. Although not immediate, this effect
illumination of 1,000 Wm− 2–1 sun), calibrated with a reference silicon was gradually increased as days passed by. Fig. 4 shows pictures of the
solar cell from PV Measurements before device measurements. Current samples taken after 24 h of curing. Nevertheless, it did not affect the
density-Voltage (J-V) measurements were performed in reverse and performance of the cells. A possible explanation could be that the

4
G.A. Soares et al. Solar Energy 273 (2024) 112496

Fig. 7. (a) Microscope image of analyzed region showing measurements taken in the perovskite with Ag region, intermittent region, and silver-unprotected region
(b) Raman spectrum illustrating the surface composition of encapsulated perovskite device. PL mapping of laminated devices (c) 2D and (d) 3D images. The bar
indicates the intensity of emission when excited at 438 nm..

Fig. 8. Testing of adhesive lateral penetration. In the standard method, (a) the adhesive was placed on top of the coated substrate and (b) after curing the color is
removed from the areas not protected by the silver electrode. (c) A laser cut was then made around and slicing the cell in order to evaluate if the contact of the
adhesive within the layers below BCP, prior to curing causes lateral degradation.

evaporated silver top contact protected the underneath layers from the indicates that a chemical reaction occurred between the epoxy adhesive
adhesive. As for the acrylic adhesives, visual and electrical character­ and the perovskite and also that the Ag layer effectively avoided the
istics remained constant after the encapsulation process. However, the contact of the adhesive with the perovskite/PCBM/BCP films, prevent­
sample with Acrylic B did not seem to be properly cured, since it ing the degradation. Absorbance spectra taken in the area between the
remained sticky on the edges were there was excess of glue. The appli­ cells immediately after the encapsulation and 24 h and 96 h after con­
cation of additional UV exposure or heat did not improve this condition, firms the gradual vanishing of the photo-absorber, as presented Fig. 6.
which might indicate as well some chemical incompatibility. We conducted X-ray diffraction (XRD) and Raman spectroscopy an­
It has been previously reported that outgassed vapors of UV-curable alyses to assess the encapsulated samples. Given the observable change
epoxies could damage the perovskite during the UV-curing [20]. Optical in film color, we conducted a thorough investigation to elucidate the
microscope images of edges of the cells encapsulated with the epoxy underlying causes.
resins were taken in order to investigate any degradation under the cells The XRD measurements aimed to provide a deeper understanding of
due to glue penetration either through the top layer or laterally. As it perovskite crystal growth when utilizing the blade coating technique. In
could be observed from Fig. 5, the perovskite characteristic color dis­ the presented Figure S3, XRD patterns of the perovskite films deposited
appeared from the area where there was no evaporated Ag. This on ITO/PEDOT:PSS substrates are illustrated. These patterns revealed

5
G.A. Soares et al. Solar Energy 273 (2024) 112496

and diligence applied, we encountered a persistent fluorescence signal

that exhibited substantial overlap with the desired structural features.
This phenomenon presented a unique challenge in our quest for an un­
obstructed understanding of the underlying material properties. The
results reflect the intricacies of our investigation and emphasize the
necessity of careful consideration in the study of advanced materials and
encapsulation techniques. The findings underline the potential presence
of fluorescence-related interferences in XRD analyses and prompt
further exploration in this domain to uncover the underlying mecha­
nisms responsible for this intriguing observation
In our quest for alternative analytical techniques, we turned to
Raman spectroscopy to investigate three distinct regions (Fig. 7a).
Raman Spectroscopy is a non-destructive chemical analysis technique
that offers in-depth insights into a chemical structure of material, phase,
polymorphism, crystallinity, and molecular interactions by examining
Fig. 9. Results of the adhesive lateral penetration test: (a) Laser cut made the interaction of light with its chemical bonds.
around the cells and over 2 cells to allow lateral contact of the adhesive with all The Raman spectrum reveals (Fig. 7b) the surface composition of
layers. (b) Image took after complete discoloration. (c) Close-up on the laser cut perovskite films deposited on ITO/PEDOT:PSS, excited at 438 nm.
over the cell on the silver side. No effects on the edge were observed. Specifically, the peak at 119 cm− 1 corresponds to the CH3NH3PbI3 thin
film, while the peaks at 76 cm− 1 and 94 cm− 1 are attributed to PbI2
distinct peaks at diffraction angles (2θ) of 14.12◦ and 28.25◦ , corre­ [32,33]. The presence of PbI2 can be attributed to the incomplete con­
sponding to the characteristic (1 1 0) and (2 2 0) planes, respectively, of version of the PbI2 precursor or it may be indicative of perovskite
methylammonium lead trihalide perovskite (CH3NH3PbI3), as previ­ degradation. In the Raman spectroscopy results, we examined the re­
ously reported in the literature[30]. Regarding XRD, we performed gions of the perovskite film that were protected with Ag layer, the edge
measurements post-lamination to monitor the impact of lamination and regions, and the unprotected areas. Notably, we observed a reduction in
assess the crystal’s stability. However, the samples exhibited an amor­ the characteristic signal of CH3NH3PbI3 in the unprotected region. On
phous structure after lamination that could not be definitively correlated the other hand, at the edges, we observed a lesser reduction in
with perovskite, interlayers, or adhesive. Consequently, XRD is not a CH3NH3PbI3 signal compared to the completely unprotected area.
suitable technique for analyzing perovskite post-lamination due to the Moreover, when evaluating the photoluminescence (Fig. 7c and d), we
presence of this amorphous structure, which interferes with the identi­ observed enhanced emission, suggesting an interaction with interlayers.
fication of the perovskite crystal structure. In our pursuit of a compre­ This can be attributed to the exciton created within the perovskite,
hensive analysis of the structural properties of our “delaminated” which, in regions of lower energy, is less likely to dissociate and more
samples, we undertook an intricate XRD examination. These samples likely to recombine. These findings highlight the influence of silver
underwent a meticulous process involving the removal of an additional protection and edge effects on the optical properties of the perovskite
glass protection layer, shedding light on the complex nature of the layer, shedding light on the exciton behavior and its impact on
encapsulation. Our XRD analysis revealed an intriguing challenge as we photoluminescence.
probed deeper into the structural characteristics. Despite the precision Further experiments were carried out to investigate the possibility of

Fig. 10. Pictures of encapsulated devices after 48 h immersed in water.

Fig. 11. (a) Top view and (b) cross-section schematics of the calcium test sample. As glass is used to encapsulate the sample, the permeation occurs only laterally
through the adhesive bulk and glass-adhesive interface.

6
G.A. Soares et al. Solar Energy 273 (2024) 112496

Fig. 12. Degradation of the calcium samples over the period of 90 days inside a climate chamber.

Fig. 13. Thermal stability of devices encapsulated with Acrylic A and Epoxy A adhesives: (a) PCE, (b) Voc, (c) Js and (d) FF. Plotted data represent the average value
of the 8 cells in each sample and the error bars represent the standard deviation. Measurements in reverse scan.

the lateral penetration in the cell area, below the Ag top layer. A laser 3.2. Barrier properties
scribing was done around the active area of the cells and over the active
area exposing all layers to the lateral contact of the adhesive prior to the The encapsulated samples were further tested immersed in water to
curing. However, no degradation was observed on the cell area, indi­ investigate the barrier properties of the adhesives and the rate of water
cating that the adhesive did not penetrate under the silver top electrode, penetration. Samples were evaluated visually and the results after 48 h
but rather that the reaction occurs through the direct contact with the are depicted in Fig. 10. The discoloration in the epoxy samples hindered
BCP/PCBM/Perovskite layers (Fig. 8 and Fig. 9). the assessment of the water penetration, but there was no visual
degradation in the cell area of any of the samples. On the other hand, the
lateral water ingress was notable in both acrylic ones. On the edges, the

7
G.A. Soares et al. Solar Energy 273 (2024) 112496

Fig. 14. Pictures of representative samples before (top) and after (bottom) the thermal test. (a) and (c) show the acrylic samples before and after the thermal test,
respectively. (b) and (d) show the epoxy sample at the same conditions.

working with organic photovoltaics (OPV) at Oninn.

The calcium test has been commonly used to test moisture and ox­
ygen barriers, and recently has been demonstrated the relationship be­
tween optical calcium test and PSC performances [37] therefore, an
optical calcium test was then performed to confirm the previous results
and to understand the better barrier property of the epoxy-based ad­
hesive compare to the acrylic-based one. In its metallic form, calcium
has a silver color, but when in contact with moisture, it is converted into
calcium hydroxide, Ca(OH)2, which is white. Through this color change,
it is possible to visually assess where the penetration occurred as well as
the rate, when monitored over time [38,39]. The schematics of the
samples can be seen in Fig. 11.
The samples were prepared and encapsulated in the glove box to
avoid degradation. After curing, they were placed inside an environ­
mental chamber, constantly set at 65 ◦ C and 85 % RH. Pictures of the
samples, encapsulated with the acrylic and the epoxy adhesives were
taken periodically to follow the degradation, as presented in Fig. 12.
Along three months of test, the sample with the epoxy adhesive kept
its color, with shape and dimensions barely changed. On the other hand,
Fig. 15. Results before and after encapsulation of devices with modified HTL. the acrylic sample gradually turned from a square to a round shape of
Solid lines represent measurements in reverse scan, whilst dashed lines repre­ smaller dimensions, indicating that the water penetrated laterally the
sent forward scan measurements. sample and reacted with calcium. According to [40], permeability of
barrier is proportional to calcium area reduction, therefore, it was
characteristic brownish color of the perovskite was turned into yellow, concluded that Acrylic A had lower barrier properties when compared to
as a result of CH3NH3PbI3 decomposition into solid PbI2 and MAI+ due Epoxy A since from initial day (0 day) until the end of the test (90 days)
to the contact with water [34–36]. This effect seemed to be more pro­ the calcium area decreased ~ 70 % and ~ 5 %, respectively. This ex­
nounced in Acrylic B, which could be associated with its incomplete plains the result of the immersion test previously shown and is in
cure, as previously observed. accordance with permeation values (WVTR) reported in the data sheet
For the continuity of the tests, it was decided to use only one version of the materials (SI Table S1).
of each type of adhesive. Acrylic B was discarded due to its curing issue,
and Epoxy A was chosen over B due to the experience accumulated

8
G.A. Soares et al. Solar Energy 273 (2024) 112496

Fig. 16. Thermal stability at 85 ◦ C of PEDOT:PSS and NiO devices encapsulated with the epoxy adhesive: (a) PCE, (b) Voc, (c) Js and (d) FF. Plotted data represent
the average value of the 8 cells in each sample and the error bars represent the standard deviation. Measurements in reverse scan.

thermal stability test.

Devices with the epoxy adhesive showed better stability under
thermal stress. A 20 % drop in performance was observed after the first
100 h of test as a result of the drops in Voc and Jsc. As the test continued,
the degradation followed a slower rate, with more significant changes of
Jsc and FF in the end. The acrylic samples, on the other hand, underwent
a more sudden degradation, with highlights to Voc and Jsc. By analyzing
the pictures before and after the test, depicted in Fig. 14, it is possible to
see a severe degradation in the acrylic sample, which seemed to have
developed inwards, in agreement with the calcium samples. This could
indicate that degradation, in this case, could have been more related to
the lateral penetration of humidity and oxygen, accelerated by heat. In
the picture of the epoxy sample is notable that the surrounding areas
became more transparent, while the cells were mostly preserved.
A significant drop in FF and Jsc was observed for one of the epoxy
samples after 700 h. This could be justified by the failure of two cells
(second and third from the right side of Fig. 13(d)), where is possible to
see the beginning of degradation due to lateral penetration.

3.4. HTL modification

The good results presented by the epoxy adhesive allowed it to be

used to evaluate and compare the stability of devices fabricated with
different HTLs: PEDOT:PSS and NiO, which are the most used HTL in p-i-
Fig. 17. Encapsulated flexible perovskite device. The color of the surrounding n PSCs. Results before and after encapsulation are depicted in Fig. 15
areas of the cells, not protected by the top electrode, is vanished, producing a and show no significant effect of the encapsulation over the electrical
semi-transparent device. performance of the devices. For the NiO samples, we observed less
dispersed data after encapsulation and a slight increase in the mean
3.3. Thermal stability value in the reverse and forward scan, which could be attributed to a
light soaking effect. Besides that, the encapsulation may have contrib­
PSCs encapsulate with Acrylic A and Epoxy A were tested under uted to a more stable environment, reducing the influence of external
constant heat at 85 ◦ C, following the ISOS-D-2 protocol [41]. No sig­ factors and providing more consistent conditions for the NiO-based
nificance change in performance was observed before and after encap­ devices. This enhanced stability, in turn, could lead to the observed
sulation, the results are presented in Figure S4 in Support Information. increase in the mean efficiency after encapsulation. Under heat, how­
Fig. 13 shows the evolution of the electrical parameters during the ever, despite higher initial efficiency and slower degradation up to 800
h, the device with PEDOT:PSS degraded faster, following the drops of FF

9
G.A. Soares et al. Solar Energy 273 (2024) 112496

Fig. 18. (a) PCE before and after lamination of cells on a flexible substrate using Epoxy A. (b) J-V curves of the best cell before and after lamination. Solid lines
represent measurements in reverse scan, whilst dashed lines represent forward scan measurements.

and Jsc. Fig. 16 shows the evolution of the main electrical parameters CRediT authorship contribution statement
during the test.
As already described in the literature [42,43], the nature of PEDOT: Gabriela A. Soares: Conceptualization, Methodology, Validation,
PSS is an important factor for the degradation of the perovskite device. Formal analysis, Writing – original draft. Izabela S. Bicalho: Validation,
As a hygroscopic and acidic material, it can promote the decomposition Resources. Sergio Castro-Hermosa: Conceptualization, Resources,
of the perovskite through the absorbed water as well as accelerate the Formal analysis. Luiza de Q. Corrêa: Validation, Resources. Bárbara H.
corrosion of the ITO electrode. On the other hand, NiO has a stable S. Miranda: Validation, Resources. Adriano dos S. Marques: Valida­
crystal structure and do not present corrosive characteristics, giving the tion, Resources. Silvia L. Fernandes: Validation, Resources. Thainá
devices higher thermal stability [44]. Cunha: Validation, Resources. Vinícius V. de Freitas: Validation, Re­
In order to further confirm these results, samples with a more stable sources. Rodrigo de Q. Vilaça: Validation, Resources, Funding acqui­
structure, using NiO as HTL and a CsFA-based perovskite, deposited on a sition. Luana Wouk: Conceptualization, Methodology, Validation,
flexible substrate, were laminated using Epoxy A and a flexible barrier Formal analysis, Writing – review & editing. Diego Bagnis: Conceptu­
film. Further information on these devices can be found in SI. A picture alization, Project administration, Funding acquisition, Supervision.
of one device is shown in Fig. 17. The transparency between the cells is
caused by the same phenomena observed with the glass samples. Declaration of competing interest
As it can be seen in Fig. 18, the encapsulation process resulted in a
loss of less than 10 % in performance of the cells, which presented a The authors declare that they have no known competing financial
small reduction of Voc, but an increase in current density, as demon­ interests or personal relationships that could have appeared to influence
strated by the curve. This result, besides confirming the possibility to use the work reported in this paper.
the epoxy adhesive with a different perovskite structure, also demon­
strates its applicability on flexible devices – a close reality for perovskite Acknowledgments
solar cells.
The authors acknowledge Oninn for the support and the Organic &
4. Conclusion Printed Electronics R&D team. We acknowledge Petróleo Brasileiro S.A.
(PETROBRAS) under the project PD-00553-005/2018 for funding.
The comparison of the encapsulation adhesives showed that the
epoxy resins had better barrier properties than the acrylic ones. Moisture Appendix A. Supplementary data
and oxygen permeabilities of adhesives were studied using an optical
calcium test which shown that after 90 days at 65 ◦ C and 85 % RH, the Supplementary data to this article can be found online at https://doi.
calcium area of devices encapsulated with epoxy resin decreased only 5 org/10.1016/j.solener.2024.112496.
%, being much more efficient than those with the acrylic resin (calcium
area decreased 70 %). Also, the results presented in this study showed References
that it is possible to use the epoxy adhesive to encapsulate and evaluate
the stability of perovskite based solar cell devices in ambient air and [1] H. Min, D.Y. Lee, J. Kim, G. Kim, K.S. Lee, J. Kim, M.J. Paik, Y.K. Kim, K.S. Kim, M.
G. Kim, T.J. Shin, S. Il Seok, Perovskite solar cells with atomically coherent
under thermal stress. The comparison of HTLs between PEDOT:PSS and
interlayers on SnO2 electrodes, Nature 598 (2021) 444–450, https://doi.org/
NiO (at 85 ◦ C) confirmed that the oxide-based device was stable for more 10.1038/s41586-021-03964-8.
than 5000 h after the initial drop in performance, as expected. [2] M. Green, E. Dunlop, J. Hohl-Ebinger, M. Yoshita, N. Kopidakis, X. Hao, Solar cell
Nevertheless, epoxy resins chemically react when in contact with the efficiency tables (version 57), Prog. Photovolt. Res. Appl. 29 (2021) 3–15, https://
doi.org/10.1002/pip.3371.
perovskite films. A solid and compact top electrode, as in the case of [3] Y. Cheng, L. Ding, Pushing commercialization of perovskite solar cells by
evaporated silver, can offer a long-term protection to the layers below it, improving their intrinsic stability, Energy Environ. Sci. 14 (2021) 3233–3255,
as it was showed. However, it is recommended and necessary that future https://doi.org/10.1039/d1ee00493j.
[4] S. Mazumdar, Y. Zhao, X. Zhang, Stability of perovskite solar cells: Degradation
studies focus on finding or developing an inert material (respect to mechanisms and remedies, Front. Electron. 2 (2021) 1–34, https://doi.org/
perovskite) to be used as a protective layer, which would enable the 10.3389/felec.2021.712785.
encapsulation in cases where the electrode does not cover the surface [5] P. Yadav, D. Prochowicz, E.A. Alharbi, S.M. Zakeeruddin, M. Grätzel, Intrinsic and
interfacial kinetics of perovskite solar cells under photo and bias-induced
entirely, as in the grid design, or where other printing methods that do degradation and recovery, J. Mater. Chem. C Mater. 5 (2017) 7799–7805, https://
not result in such compact layer are used. doi.org/10.1039/c7tc02652h.
[6] S. Zhang, G. Han, Intrinsic and environmental stability issues of perovskite
photovoltaics, Prog. Energy 2 (2020) 022002, https://doi.org/10.1088/2516-
1083/ab70d9.
[7] S.P. Dunfield, L. Bliss, F. Zhang, J.M. Luther, K. Zhu, M.F.A.M. van Hest, M.
O. Reese, J.J. Berry, From defects to degradation: A mechanistic understanding of

10
G.A. Soares et al. Solar Energy 273 (2024) 112496

degradation in perovskite solar cell devices and modules, Adv. Energy Mater. 10 [26] F. Corsini, G. Griffini, Recent progress in encapsulation strategies to enhance the
(2020) 1–35, https://doi.org/10.1002/aenm.201904054. stability of organometal halide perovskite solar cells, J. Phys. Energy 2 (2020),
[8] Z. Fu, M. Xu, Y. Sheng, Z. Yan, J. Meng, C. Tong, D. Li, Z. Wan, Y. Ming, A. Mei, https://doi.org/10.1088/2515-7655/ab8774.
Y. Hu, Y. Rong, H. Han, Encapsulation of printable mesoscopic perovskite solar [27] J. Li, R. Xia, W. Qi, X. Zhou, J. Cheng, Y. Chen, G. Hou, Y. Ding, Y. Li, Y. Zhao,
cells enables high temperature and long-term outdoor stability, Adv. Funct. Mater. X. Zhang, Encapsulation of perovskite solar cells for enhanced stability: Structures,
29 (2019), https://doi.org/10.1002/adfm.201809129. materials and characterization, J. Power Sources 485 (2021) 229313, https://doi.
[9] K.A. Bush, A.F. Palmstrom, Z.J. Yu, M. Boccard, R. Cheacharoen, J.P. Mailoa, D. org/10.1016/j.jpowsour.2020.229313.
P. McMeekin, R.L.Z. Hoye, C.D. Bailie, T. Leijtens, I.M. Peters, M.C. Minichetti, [28] A. Uddin, M.B. Upama, H. Yi, L. Duan, Encapsulation of organic and perovskite
N. Rolston, R. Prasanna, S. Sofia, D. Harwood, W. Ma, F. Moghadam, H.J. Snaith, solar cells: A review, Coatings 9 (2019), https://doi.org/10.3390/
T. Buonassisi, Z.C. Holman, S.F. Bent, M.D. McGehee, 23.6%-efficient monolithic coatings9020065.
perovskite/silicon tandem solar cells with improved stability, Nat. Energy 2 (2017) [29] R.K. Raman, S.A. Gurusamy Thangavelu, S. Venkataraj, A. Krishnamoorthy,
1–7, https://doi.org/10.1038/nenergy.2017.9. Materials, methods and strategies for encapsulation of perovskite solar cells: From
[10] R. Cheacharoen, N. Rolston, D. Harwood, K.A. Bush, R.H. Dauskardt, M. past to present, Renew. Sustain. Energy Rev. 151 (2021) 111608, https://doi.org/
D. McGehee, Design and understanding of encapsulated perovskite solar cells to 10.1016/j.rser.2021.111608.
withstand temperature cycling, Energy Environ. Sci. 11 (2018) 144–150, https:// [30] S. Castro-Hermosa, L. Wouk, I.S. Bicalho, L. de Queiroz Corrêa, B. de Jong, L. Cinà,
doi.org/10.1039/c7ee02564e. T.M. Brown, D. Bagnis, Efficient fully blade-coated perovskite solar cells in air with
[11] R. Cheacharoen, C.C. Boyd, G.F. Burkhard, T. Leijtens, J.A. Raiford, K.A. Bush, S. nanometer-thick bathocuproine buffer layer, Nano Res. 14 (2021) 1034–1042,
F. Bent, M.D. McGehee, Encapsulating perovskite solar cells to withstand damp https://doi.org/10.1007/s12274-020-3147-4.
heat and thermal cycling, Sustain. Energy Fuels 2 (2018) 2398–2406, https://doi. [31] F. Yu, Y. Liang, X. Sun, G. Huang, C. Xu, Modelling solar cells’ S-shaped I-V
org/10.1039/c8se00250a. characteristics with an analytical solution to lumped-parameter equivalent circuit
[12] L. Shi, T.L. Young, J. Kim, Y. Sheng, L. Wang, Y. Chen, Z. Feng, M.J. Keevers, model, Sol. Energy 202 (2020) 498–506, https://doi.org/10.1016/j.
X. Hao, P.J. Verlinden, M.A. Green, A.W.Y. Ho-Baillie, Accelerated lifetime testing solener.2020.03.090.
of Organic− Inorganic perovskite solar encapsulated by polyisobutylene.pdf, ACS [32] J. Yang, A. Xiao, L. Xie, K. Liao, X. Deng, C. Li, A. Wang, Y. Xiang, T. Li, F. Hao,
Appl. Mater. Interfaces 9 (2017) 25073–25081. Precise control of PbI2 excess into grain boundary for efficacious charge extraction
[13] E. Ramasamy, V. Karthikeyan, K. Rameshkumar, G. Veerappan, Glass-to-glass in off-stoichiometric perovskite solar cells, Electrochim. Acta 338 (2020) 135697,
encapsulation with ultraviolet light curable epoxy edge sealing for stable https://doi.org/10.1016/j.electacta.2020.135697.
perovskite solar cells, Mater. Lett. 250 (2019) 51–54, https://doi.org/10.1016/j. [33] Y. Yan, W.-J. Yin, T. Shi, W. Meng, C. Feng, Defect physics of CH3NH3PbX3 (X = I,
matlet.2019.04.082. Br, Cl) perovskites, in: N. Park, M. Grätzel, T. Miyasaka (Eds.), Organic-Inorganic
[14] B. McKenna, J.R. Troughton, T.M. Watson, R.C. Evans, Enhancing the stability of Halide Perovskite Photovoltaics, Springer International Publishing, Cham, 2016,
organolead halide perovskite films through polymer encapsulation, RSC Adv. 7 pp. 79–105, https://doi.org/10.1007/978-3-319-35114-8_4.
(2017) 32942–32951, https://doi.org/10.1039/c7ra06002e. [34] H. Xu, A brief review on the moisture stability for perovskite solar cells, IOP Conf.
[15] Y.H. Soo, S.A. Ng, Y.H. Wong, C.Y. Ng, Thermal stability enhancement of Ser. Earth Environ. Sci. 585 (2020), https://doi.org/10.1088/1755-1315/585/1/
perovskite MAPbI3 film at high temperature (150 ◦ C) by PMMA encapsulation, 012027.
J. Mater. Sci. Mater. Electron. 32 (2021) 14885–14900, https://doi.org/10.1007/ [35] Z. Song, A. Abate, S.C. Watthage, G.K. Liyanage, A.B. Phillips, U. Steiner,
s10854-021-06041-y. M. Graetzel, M.J. Heben, Perovskite solar cell stability in humid air: Partially
[16] Z. Liu, B. Sun, T. Shi, Z. Tang, G. Liao, Enhanced photovoltaic performance and reversible phase transitions in the PbI2-CH3NH3I-H2O system, Adv. Energy Mater. 6
stability of carbon counter electrode based perovskite solar cells encapsulated by (2016), https://doi.org/10.1002/aenm.201600846.
PDMS, J. Mater. Chem. A Mater. 4 (2016) 10700–10709, https://doi.org/10.1039/ [36] S. Kundu, T.L. Kelly, In situ studies of the degradation mechanisms of perovskite
c6ta02851a. solar cells, EcoMat 2 (2020) 1–22, https://doi.org/10.1002/eom2.12025.
[17] A. Smith, T.J. Routledge, B.A. Alqurashy, B.G. Freestone, A.J. Parnell, N. Vaenas, [37] S. Castro-Hermosa, M. Top, J. Dagar, J. Fahlteich, T.M. Brown, Quantifying
M.O.A. Alawad, A. Iraqi, C. Rodenburg, D.G. Lidzey, High-performance multilayer performance of permeation barrier—Encapsulation systems for flexible and glass-
encapsulation for perovskite photovoltaics, Adv. Energy Mater. 8 (2018), https:// based electronics and their application to perovskite solar cells, Adv. Electron.
doi.org/10.1002/aenm.201801234. Mater. (2019) 1–11, https://doi.org/10.1002/aelm.201800978.
[18] F. Matteocci, L. Cinà, E. Lamanna, S. Cacovich, G. Divitini, P.A. Midgley, C. Ducati, [38] P. Boldrighini, A. Fauveau, S. Thérias, J.L. Gardette, M. Hidalgo, S. Cros, Optical
A. Di Carlo, Encapsulation for long-term stability enhancement of perovskite solar calcium test for measurement of multiple permeation pathways in flexible organic
cells, Nano Energy 30 (2016) 162–172, https://doi.org/10.1016/j. optoelectronic encapsulation, Rev. Sci. Instrum. 90 (2019), https://doi.org/
nanoen.2016.09.041. 10.1063/1.5021182.
[19] Q. Dong, F. Liu, M.K. Wong, H.W. Tam, A.B. Djurišić, A. Ng, C. Surya, W.K. Chan, [39] S. Schubert, H. Klumbies, L. Müller-Meskamp, K. Leo, Electrical calcium test for
A.M.C. Ng, Encapsulation of perovskite solar cells for high humidity conditions, moisture barrier evaluation for organic devices, Rev. Sci. Instrum. 82 (2011),
ChemSusChem 9 (2016) 2518, https://doi.org/10.1002/cssc.201601090. https://doi.org/10.1063/1.3633956.
[20] S. Ma, Y. Bai, H. Wang, H. Zai, J. Wu, L. Li, S. Xiang, N. Liu, L. Liu, C. Zhu, G. Liu, [40] S. Hergert, M. Linkor, M. Korny, N. Fruehauf, Process development and accurate
X. Niu, H. Chen, H. Zhou, Y. Li, Q. Chen, 1000 h operational lifetime perovskite low-cost characterization for OLED sealants by using a calcium test, J Soc. Inf.
solar cells by ambient melting encapsulation, Adv. Energy Mater. 10 (2020), Disp. 15 (2007) 421, https://doi.org/10.1889/1.2749328.
https://doi.org/10.1002/aenm.201902472. [41] M.O. Reese, S.A. Gevorgyan, M. Jørgensen, E. Bundgaard, S.R. Kurtz, D.S. Ginley,
[21] F.J. Ramos, T. Maindron, S. Béchu, A. Rebai, M. Frégnaux, M. Bouttemy, D.C. Olson, M.T. Lloyd, P. Morvillo, E.A. Katz, A. Elschner, O. Haillant, T.
J. Rousset, P. Schulz, N. Schneider, Versatile perovskite solar cell encapsulation by R. Currier, V. Shrotriya, M. Hermenau, M. Riede, K.R. Kirov, G. Trimmel, T. Rath,
low-temperature ALD-Al2O3 with long-term stability improvement, Sustain. Energy O. Inganäs, F. Zhang, M. Andersson, K. Tvingstedt, M. Lira-Cantu, D. Laird,
Fuels 2 (2018) 2468–2479, https://doi.org/10.1039/c8se00282g. C. McGuiness, S. Gowrisanker, M. Pannone, M. Xiao, J. Hauch, R. Steim, D.
[22] H. Wang, Y. Zhao, Z. Wang, Y. Liu, Z. Zhao, G. Xu, T.H. Han, J.W. Lee, C. Chen, M. Delongchamp, R. Rösch, H. Hoppe, N. Espinosa, A. Urbina, G. Yaman-Uzunoglu,
D. Bao, Y. Huang, Y. Duan, Y. Yang, Hermetic seal for perovskite solar cells: An J.B. Bonekamp, A.J.J.M. Van Breemen, C. Girotto, E. Voroshazi, F.C. Krebs,
improved plasma enhanced atomic layer deposition encapsulation, Nano Energy Consensus stability testing protocols for organic photovoltaic materials and
69 (2020) 104375, https://doi.org/10.1016/j.nanoen.2019.104375. devices, Sol. Energy Mater. Sol. Cells 95 (2011) 1253–1267, https://doi.org/
[23] V. Zardetto, B.L. Williams, A. Perrotta, F. Di Giacomo, M.A. Verheijen, 10.1016/j.solmat.2011.01.036.
R. Andriessen, W.M.M. Kessels, M. Creatore, Atomic layer deposition for perovskite [42] C. Redondo-Obispo, T.S. Ripolles, S. Cortijo-Campos, A.L. Álvarez, E. Climent-
solar cells: Research status, opportunities and challenges, Sustain. Energy Fuels 1 Pascual, A. de Andrés, C. Coya, Enhanced stability and efficiency in inverted
(2017) 30–55, https://doi.org/10.1039/c6se00076b. perovskite solar cells through graphene doping of PEDOT:PSS hole transport layer,
[24] S. Emami, J. Martins, D. Ivanou, A. Mendes, Advanced hermetic encapsulation of Mater Des. 191 (2020), https://doi.org/10.1016/j.matdes.2020.108587.
perovskite solar cells: The route to commercialization, J. Mater. Chem. A Mater. 8 [43] L. Cai, F. Zhu, Toward efficient and stable operation of perovskite solar cells:
(2020) 2654–2662, https://doi.org/10.1039/c9ta11907h. Impact of sputtered metal oxide interlayers, Nano Select. 2 (2021) 1417–1436,
[25] S. Emami, J. Martins, R. Madureira, D. Hernandez, G. Bernardo, J. Mendes, https://doi.org/10.1002/nano.202000290.
A. Mendes, Development of hermetic glass frit encapsulation for perovskite solar [44] L. Xu, X. Chen, J. Jin, W. Liu, B. Dong, X. Bai, H. Song, P. Reiss, Inverted perovskite
cells, J. Phys. D Appl. Phys. (2018), https://doi.org/10.1016/j. solar cells employing doped NiO hole transport layers: A review, Nano Energy 63
fishres.2006.06.005. (2019) 103860, https://doi.org/10.1016/j.nanoen.2019.103860.

11