In the collective imagination, the vision of a sustainable future, in which the growing demands for space and energy of an ever-expanding human population are fulfilled while respecting the environment, is often associated with ethereal metropolitan skylines of self-powered glass buildings. Owing to advanced materials concepts, emerging energy technologies1 and for-ward-thinking public policies2, this ideal scenario might be less remote than it seems. The transition to fully energetically sustainable architecture is, in fact, already in progress, and so-called net zero-energy buildings (NZEBs), whose yearly energy consumption is nearly fully counterbalanced by renewable energy generated on site, are rapidly becoming valuable alternatives to conventional dwellings while also providing rapid returns on investment. For individual houses and small buildings, energy microgeneration technologies, such as rooftop photovoltaics (PV), micro-turbines,

building) in the form of PV skins or glazing units (BOX 1). Because BIPV systems are effectively part of the urban landscape, the key element for their public acceptance and, ultimately, their diffusion is their ‘invisible’ integration without affecting the aesthetics of a building or reducing the quality of life of its inhabitants.

Luminescent solar concentrators (LSCs) could facilitate the green architecture revolution by enabling the realization of semi-transparent PV glazing systems, which could potentially convert the façades of urban buildings into distributed electrical power generators3–7 (FIG. 1a). LSCs were first proposed in 1976 as cost-effective alternatives to silicon solar cells8,9. They offered the possibility of exploiting solar radiation by use of large-area devices that needed a minimal amount of PV material, which was prohibitively expensive at that time. To date, as a result of the remarkable drop in the retail cost of silicon PV modules (~70% over the past decade), the substitution of silicon panels for rooftop applications or solar farming is no longer a priority. However, owing to their structure and all-optical functioning mechanism, LSCs offer a unique opportunity for ‘invisible’ architectural integration of solar energy devices in cities, elegantly circumventing the aesthetic and functional limitations of opaque and electrode-based semi- transparent PV modules10 that hinder their public acceptance and market penetration.

In its most common embodiment, an LSC consists of a polymeric or glassy optical waveguide doped or coated with chromophores11,12 (FIG. 1a). Direct or diffused sunlight penetrates the waveguide and is absorbed by the chromophores, which re-emit it at longer wavelength. The emitted light is trapped inside the waveguide and guided by total internal reflection to the edges, where small index-matched PV cells convert it into electricity11,13–16. This all-optical operation, together with the design and fabrication versatility of LSCs, has several advantages over conventional PV systems integrated into building façades. In particular, LSCs are less sensitive than silicon PV modules to their orientation angle, and the indirect illumination of the perimeter cells by the waveguide makes LSCs nearly

geothermal heat pumps and solar thermal collectors, are available to meet electricity, heating and cooling needs. High-performing insulation strategies, in turn, mitigate energy dissipation and consumption. However, the realization of NZEBs is more complex in highly urbanized environments because the cost of land for ground PV installation is prohibitively high and the rooftop space is insufficient to accommodate the PV modules necessary for meeting the electrical requirements of tall buildings (roughly 7 m2 is currently needed per kilowatt of peak power). Moreover, in the near future, a further dramatic increase in electrical demands is expected as a consequence of the widespread use of fully electrical vehicles. For this reason, increasing efforts are being dedicated to the realization of building-in-tegrated PV (BIPV) technologies, in which PV elements become an integral part of the building envelope (the outer shell physically separating the interior and exterior of a

Luminescent solar concentrators for building-integrated photovoltaicsFrancesco Meinardi, Francesco Bruni and Sergio Brovelli

Abstract | The transition to fully energetically sustainable architecture through the realization of so-called net zero-energy buildings is currently in progress in areas with low population density. However, this is not yet true in cities, where the cost of land for the installation of ground photovoltaic (PV) is prohibitively high and the rooftop space is too scarce to accommodate the PV modules necessary for sustaining the electrical requirements of tall buildings. Thus, new technologies are being investigated to integrate solar-harvesting devices into building façades in the form of PV windows or envelope elements. Luminescent solar concentrators (LSCs) are the most promising technology for semi- transparent, electrodeless PV glazing systems that can be integrated ‘invisibly’ into the built environment without detrimental effects to the aesthetics of the building or the quality of life of the inhabitants. After 40 years of research, recent breakthroughs in the realization of reabsorption-free emitters with broadband absorption have boosted the performance of LSCs to such a degree that they might be commercialized in the near future. In this Perspective, we explore the successful strategies that have allowed this change of pace, examining and comparing the different types of chromophores and waveguide materials, and discuss the issues that remain to be investigated for further progress.

PERSPECTIVES

NATURE REVIEWS | MATERIALS VOLUME 2 | ARTICLE NUMBER 17072 | 1

Dol

lars

in b

illio

ns

NZEBsPrestige buildingsCommercial and government buildingsResidential buildingsIndustrial buildings

Growth of the BIPV glazing sector from 2015 to 2022

20150

1

2

3

4

5

6

7

2016 2017 2018 2019

Year2020 2021 2022

unaffected by efficiency losses and electrical stresses caused by shadowing effects that occur in bulk and thin-film PVs. Crucially, the shape, transparency, colour and flexibility of LSCs can be controlled to a great extent6,17 (FIG. 1b), and solar energy can be collected using electrodeless semi- transparent waveguides with essentially no aesthetic impact7,18,19. This makes LSCs ideally suitable for a number of BIPV applications, such as windows, coverages18 and sound barriers20 (FIG. 1c), and LSCs could even become a tool for architects to enhance the aesthetical value of a building3,21. Furthermore, the insertion of the LSC panel into triple-insulated glazing units in substitution of the energetically inert internal glass panel (FIG. 1d) simultaneously

chromophores and waveguide matrices and highlight future strategies for developing LSC-based technologies for real-world BIPV applications.

Optimization of the chromophoresTo fully exploit the potential of LSCs in BIPV, it is necessary to simultaneously maximize the solar-harvesting capability of a given chromophore, which determines the fraction of solar light that can be collected and converted into electricity, and the power efficiency of the device, which requires the implementation of specific strategies for suppressing detrimental optical losses. Among the loss processes (BOX 2), the most important is the reabsorption of the guided photolumi-nescence by the chromophores themselves, which is due to the large overlap between the absorption and photo luminescence spectra of conventional emitters.

Types of chromophores. The suppression of reabsorption has been addressed using various strategies, which led to the realization of potentially suitable LSC chromophores that can be divided into two categories based on their photoluminescence mechanisms (FIG. 2a). In type-A chromophores, photo-luminescence occurs from the light-absorbing state and reabsorption losses are mitigated by the combination of a relatively large Stokes shift (the separation between the absorption and emission peak) and sharp absorption and emission bands, which results in reduced spectral overlap. Examples of type-A emitters are organic molecular dyes, such as 4-dicyanomethyl-6-dimethylamin-ostiryl-4H-pyran (DCM)22, Lumogen Red23, CRS040 Yellow (from Radiant Colour)4 and perylene derivatives24,25. Colloidal nanocrystals of metal chalcogenides, such as CdSe or PbS, could, in principle, belong to the type-A category, but their Stokes shift is too small to achieve good performance given the width of their spectral features5,26–28. By contrast, in type-B chromophores, the absorption and emission processes are decoupled. Light harvesting occurs in a state with a large broadband absorption coefficient (α1), whereas emission takes place from a second state with a negligible absorption coefficient (α2) resonant with the emission band; reabsorption losses are effectively suppressed for α1>>α2. The optical properties of representative chromophores of the two categories are presented in FIG. 2b; a comprehensive survey of LSC emitters can be found in REF. 11.

The first type-B chromophores were organic complexes of lanthanide ions,

ensures optimal heat insulation, sufficient natural lighting for internal illumination and reduced overheating by excessive solar irradiation.

Despite this great potential, the widespread use of LSCs has been hindered by the difficulty in synthesizing suitable chromophores that fulfil all the requisites for efficient solar collection and concentration. However, owing to the development of new design strategies for high-performing LSC chromophores, the most critical bottleneck processes have been overcome, leading to a renaissance of the field. In the following, we review the most important recent successes in LSC research, discuss the remaining challenges in the design of optimized

Box 1 | Market forecast for building-integrated photovoltaics at a glance

Buildings account for 40% of the energy consumed in Europe every year. The reduction of energy

consumption and the increase of energy production from renewable sources in the building

sector are key to reducing greenhouse gas emissions. In 2015, the European Commission

indicated in the Strategic Energy Technology Plan a set of ten pivotal research and innovation

goals that embody the European priorities on low-carbon technologies. One of these goals is

enabling the mass realization of net zero-energy buildings (NZEBs) using building-integrated

photovoltaics (BIPV) through the establishment of collaborative innovation efforts between the

photovoltaic (PV) sector and key players from the building industry. Thus, the BIPV technology is

one of the key elements of the European climate policy, and it offers increasing opportunities in

terms of investment and market share. The annual installed production capacity of BIPV

worldwide is projected to exceed 11 GW by 2020, with a compound annual growth rate (CAGR)

of 30%, driven by the increasing focus on renewable energy and by the green building movement

in the construction sector. The global market (according to a Global Industry Analysts, Inc. report)

is dominated by Europe, which has a share of approximately 40% — Europe benefits from the

highest consumer willingness to adopt green practices and to comply with energy building

regulations — followed by North America with 27% of annual installations. Asia Pacific is

expected to experience the fastest growth, with a CAGR of 48% over the next 5 years. In Europe, the BIPV market includes approximately 200 commercially available products. New products that

incorporate PV modules into actual building materials, such as curtain walls, windows and roofing

shingles, are available from a variety of developers in the BIPV supply chain. The economic

advantage of integrated PV over more common non-integrated systems is that the initial cost can

be offset by reducing the amount spent on building materials and labour for those parts of the

building that can be replaced by BIPV modules. These advantages make BIPV one of the

fastest-growing segments of the PV industry. According to an n-tech Research report published in

February 2017, the total BIPV market is expected to grow from about US$3 billion in 2015 to over $9 billion in 2019 and surge to $26 billion by 2022. The BIPV glazing sector, as illustrated in the figure, will represent roughly $2.1 billion in shipments by 2018 and $6.3 billion by 2022.

PERSPECT IVES

2 | ARTICLE NUMBER 17072 | VOLUME 2 www.nature.com/natrevmats

Nature Reviews | Materials

Triple-insulated glass unita

cb

d

PV cells

hν1

hν2

ChromophorePolymer waveguide

Luminescent solar concentrator

such as europium or ytterbium29,30. In these chromophores, the ligands harvest solar light and sensitize the photolumines-cence of the rare-earth ions, which occurs through spin-forbidden or parity-forbidden transitions with very small oscillator strength (the intrinsic probability of absorption or emission). As a result, these systems exhibit very large extinction ratios, α1/α2 ≈ 104–106. The drawback of these structures is that it is not possible to tune the atomic transition of the lanthanide, which imposes a stringent limitation on the achievable spectral coverage, as discussed below. More recently, advancements in colloidal chemistry have enabled the realization of core/shell hetero- nanocrystals, in which the luminescence from the core is sensitized by a shell composed of a material with a wider energy gap that acts as a light-harvesting antenna for the solar radiation5. Examples of these systems are spherical CdSe/CdS or PbS/CdS nanocrystals5,31, and CdSe/CdS dot-in-rods32 and tetrapods33. In principle, in these systems, the maximum achievable extinction ratio is determined only by the shell thickness. However, for increasingly large shells, light scattering increases, introducing an additional source of losses. As a consequence, for real devices34, the interplay between these two phenomena leads to α1/ α2 values of 102–103 for core sizes of ~1–2 nm. Higher extinction ratios have been obtained in doped systems such as ZnSe (REF. 35), CdSe (REF. 36) or CsPbCl3 perovskite37 nanocrystals incorporating manganese ions, in which the spin-forbidden intragap photo luminescence of the dopant is sensitized by the host particle. In this case, extinction ratios as large as α1/α2 ≈ 105 have been obtained, but, as for the lanthanide complexes, the spectral coverage is strongly limited by the fixed emission wavelength of manganese at ~590 nm. This issue has been overcome by using electronic dopants, such as copper in II–VI nanocrystals, whose emission mechanism involves an intragap state and a band-edge carrier and is thus spectrally tunable through size control and hetero-structuring36,38. Similar results have been obtained with ternary I–III–VI2 nanocrystals, such as CuInS2, AgInS2 and CuInSeS, which have an excitonic mechanism that resembles that of copper-doped nanocrystals7,39–41. In these systems, the record efficiency of 3.27% for large-area devices was achieved; these nanocrystals have the further advantage of being free of potentially toxic heavy metals such as cadmium or lead7.

Highly efficient LSCs have also been demonstrated using silicon nanocrystals

transitions in confined systems; this results in a non-vanishing oscillator strength at the band edge, which is small enough to strongly limit reabsorption of the emitted light42,43.

Spectral coverage. These examples of LSC chromophores demonstrate that several approaches are currently available for the effective suppression of reabsorption losses, even for large-area LSCs, which

with an indirect bandgap6. These structures, which also have a non-toxic composition, cannot be unequivocally classified as type-A or type-B chromophores, because their luminescence occurs from the absorbing state, like in type-A chromophores, but their Stokes shift is strictly zero, like in type-B chromophores. The photoluminescence is due to the partial relaxation of the selection rule for momentum conservation for optical

Figure 1 | Luminescent solar concentration photovoltaic windows and panels. a | Schematic rep-resentation of a photovoltaic (PV) window consisting of a triple-insulated glass unit embedding a luminescent solar concentrator (LSC) replacing the inner glass panel. The LSC component is made of a polymer waveguide incorporating reabsorption-free chromophores coupled to index-matched solar cells along the edges. b | Photograph of a flexible LSC composed of silicon nanocrystals under ultra-violet illumination. c | Examples of possible architectural integration of LSCs into public spaces, such as parking or bus stop coverages or sound barriers in highways. d | Photograph taken under ambient illumination, depicting a section of a triple-insulated glass unit incorporating a poly(methyl meth-acrylate) LSC embedding CuInS2 nanocrystals coupled to crystalline Si PV cells. h, Planck’s constant; ν, frequency. Panel b is adapted with permission from REF. 6, Macmillan Publishers Limited.

PERSPECT IVES

NATURE REVIEWS | MATERIALS VOLUME 2 | ARTICLE NUMBER 17072 | 3

Non-radiative decay Escape cone

Chromophore-related losses Matrix-related losses

Surface roughness

Aggregate Bulk defect

Emitted light

Solar light

Scattering bythe surface

Matrixabsorption

Scattering by defects

are potentially suitable for integration in buildings. However, the suppression of reabsorption losses is not sufficient to obtain high efficiency. To maximize the power output of an LSC, it is necessary to harvest the largest possible fraction of the solar spectrum, which strongly limits the use of lanthanide complexes and organic dyes. In organometallic compounds, the limited spectral coverage is due to the multistep energy-transfer mechanism from the ligand to the rare-earth element, which involves multiple thermalization processes. In organic dyes, it results from the intrinsic difficulty in shifting the absorption spectrum to the near-infrared by extending the conjugation length, because an increase in conjugation length typically leads to a progressive decrease in the dye solubility and reduces its oscillator strength, thus lowering the emission efficiency. As a result, for dyes and organometallic complexes, it is difficult to obtain a spectral coverage

solar windows7. The colour depends on both the type of chromophore embedded in the waveguide and its quantity. To compare the different chromophores in FIG. 2c, we show the Commission Internationale de l’Éclairage (CIE) chromaticity coordinates of the standard D65 illuminant (artificial solar light) transmitted by the respective LSCs in the infinite concentration limit. One common feature is the strong yellow or red colouring of LSCs embedding chromophores with a spectral coverage below 50%, which is clearly detrimental to the use of these LSCs in PV glazing systems. This strong colouring greatly reduces the colour rendering index (which is the ability of a light source to faithfully reveal the real colour of objects) of transmitted solar light and modifies the indoor-to-outdoor chromatic perception, with an effect analogous to artificially induced colour blindness7. Nevertheless, coloured devices may be suitable for some special applications, such as noise barriers20, building skins or artistic architectural elements21. Lanthanide complexes yield transparent devices because their absorption edge is in the ultraviolet spectral region; similar behaviour is observed in metal halide clusters with large Stokes shifts44. However, this transparency window spanning the whole visible spectrum imposes an intrinsic and very low limit to the maximum achievable spectral coverage and, consequently, to the LSC efficiency. By contrast, chromophores with larger spectral coverages enable colourless solar windows that reduce the intensity of solar light entering a building without altering its colour and are therefore ideally suited for BIPV applications7,41.

Comparison between the performance of type-A and type-B chromophores. Because type-A and type-B chromophores mitigate reabsorption losses through different mechanisms, to estimate the best performance achievable with each approach, it is instructive to look at the evolution of the luminescence profiles of two representative systems in LSCs with increasing sizes. It is crucial to keep in mind that a proper comparison of the reabsorption capability of different emitters can be done only for equal spectral coverages. Indeed, achieving the desired spectral coverage requires different concentrations of chromophores with different energy gaps: a higher concentration of wide-energy-gap emitters is needed to match the spectral coverage of low-gap systems with the same peak absorption coefficient. In turn, for both type-A

greater than 50% of the total solar power (FIG. 2c). A similar limitation is encountered with manganese-doped nanocrystals and core/ shell CdSe/CdS systems owing to the large bandgap of the absorbing material35,36. The issue of spectral coverage has been solved by the introduction of colloidal nanocrystals of narrow-gap semiconductors, which enable the extension of the spectral coverage to the near-infrared and, in some cases (for example, for PbSe and CuInSeS), potentially surpass the harvesting capability of crystalline silicon solar cells, therefore requiring the use of non-conventional cells for the optical-to-electrical conversion of the collected energy.

Colour of the LSC panels. Another critical issue that is directly connected to the spectral coverage of large-scale BIPV applications is the colour of the LSC panels, which ultimately determines the architectural integrability of LSC-based

Box 2 | Functioning mechanism and potential loss processes of LSCs

As already highlighted in pioneering studies13,22,51, the efficiency of a luminescent solar

concentrator (LSC, ηLSC

) is affected by several potential loss processes that depend either on the

chromophores (left panel of the figure) or on the optical quality of the waveguide (right panel).

Regarding the chromophore-dependent losses, the photoluminescence quantum yield, φPL

, must

be maximized to obtain high efficiency. The portion of light trapped inside the waveguide is

determined by the refractive index of the matrix material (n), which determines the critical angle

for total internal reflection, θc = sin−1(1/n). Luminescence approaching the waveguide/air interface

with an incidence angle θi < θc, which defines the escape cone, leaves the waveguide through its

top or bottom surface. The trapping efficiency is typically expressed as ηTR

(λ) = cosθc = (1 − 1/n2)1/2.

In the common case of a polymeric LSC with n = 1.5 embedding randomly oriented chromophores

that emit isotropically, ηTR

(λ) = 0.745 is the maximum obtainable trapping efficiency in the absence

of light-scattering defects. Studies suggest that the trapping efficiency might be enhanced by

exploiting anisotropic photoluminescence to increase the in-plane emission52,53. Light trapped

inside the waveguide can propagate towards the slab edges, unless it is reabsorbed by another

chromophore, a process that has efficiency ηRA

(λ). In this case, loss can occur by non-radiative

decay if the photoluminescence quantum yield is less than one or by re-emission inside the escape

cone. Matrix-related losses are due to optical absorption by the matrix material, which prevents

light from reaching the LSC edges, or to scattering of the propagating light (with efficiency ηS(λ))

by the slab surfaces or bulk defects (either structural defects or large chromophore aggregates),

which randomizes the propagation direction, increasing the escape cone losses. As a result of these

loss mechanisms, the maximum optical efficiency achievable by an LSC can be expressed as

ηLSC

= ∫[(1 − R(λ))(1 − e−αC(λ)d)(e−αWG(λ)L)η

TR(λ)φ

PL(1 − η

RA(λ))(1 − η

S(λ))]dλ

where R(λ) = [(n − 1)/(n + 1)]2 is the reflectivity of the waveguide (~4% for n = 1.5 at normal incidence),

αC(λ) is the absorption coefficient of the chromophore, α

WG(λ) is the absorption coefficient of the

waveguide material and d and L are the thickness of the waveguide and its length, respectively.

PERSPECT IVES

4 | ARTICLE NUMBER 17072 | VOLUME 2 www.nature.com/natrevmats

α2

α1

α1

600

Nor

mal

ized

inte

nsit

ySp

ectr

al c

over

age

(%)

DCM (1)

Lumogen Red (2)

Perylene perinone (3

)

CRS040 Yellow (4

)

Oxyminopyrazole-Yb (5

)

Eu(TTA) 3(TTPO) 2

(6)

CdSe (7)

PbS (8)

CdSe/CdS (9

)

PbS/CdS (1

0)

Mn:ZnSe (1

1)

Cu:CdSe (1

2)

CuInS 2 (1

3)

CuInSeS (14)

Si (15)

Nor

mal

ized

inte

nsit

y

1.0

0.8

0.6

0.4

0.2

0.0

1.0

0.8

0.6

0.4

0.2

0.0

100

80

60

40

20

0

850 900

a

c

b

950 1,000Wavelength (nm) Wavelength (nm)

600 850

Type-A chromophores

Absorbing state = emitting state

Type-B chromophores

Absorbing state ≠ emitting state

900 950 1,000

400 600 800

Wavelength (nm)

CuInS2

CuInSeS

Si

Cu:CdSe

PbS/CdS

Mn:ZnSe

CdSe/CdS

PbS

CdSe

Eu(TTA)3(TTPO)

2

Oxyminopyrazole-Yb

Lumogen Red

DCM

Absorption Photoluminescence

Dyes

Lanthanidecom

plexesC

olloidalnanocrystals

1,000 1,200

toes

shiengineered

colloidal nanocrystals

AbsorptionSolar emission Photoluminescence

PV GaAs

PV c-Si

PV Ge4

0.4 0.6

0.3

0.4

y

x

0.5 5 9

11 7

1

2

1236,8,10,13,14,15

and type-B emitters, the chromophore concentration intrinsically determines the overlap between the absorption and emission profiles. As a result, for a given spectral coverage, reabsorption is weaker in an LSC with diluted low-energy-gap chromophores with respect to an identical device with a high concentration of

application in LSCs. With this in mind, in FIG. 3a, we compare the intrinsic photo-luminescence spectra of Lumogen Red45 and CuInSeS nanocrystals7, where the respective emission profiles were collected at the device edges as a function of device size; the emission was simulated by Monte Carlo ray tracing (a method in which

wide-energy-gap emitters with nominally the same molar extinction coefficient at the emission wavelength. Comparing LSCs with markedly different spectral coverages artificially alters their absorption–emission spectral overlap, resulting in misleading and often overoptimistic conclusions regarding the suitability of a given chromophore for

Figure 2 | Types of LSC chromophores and spectral coverage achievable

with real emitters. a | Schematic depiction of the optical properties of two types of chromophores suitable for luminescent solar concentrators (LSCs). Type-A systems feature a step-like optical absorption spectrum (blue) with a large absorption coefficient, α1, across the majority of the solar emission spectrum (grey). Photoluminescence (red) occurs from the absorbing state with minimal overlap with the long-wavelength absorption tail. In type-B chromophores, the absorption and emission processes are decoupled. The light-harvesting state has a large broadband absorption coefficient, α1, whereas the emitting state has a negligible absorption coefficient, α2, reso-nant with the emission spectrum. For the effective suppression of reabsorp-tion losses, type-A chromophores require a sharp absorption edge and near-unity emission yield. For type-B chromophores, reabsorption losses are effectively suppressed for α1>>α2. b | Optical absorption and photo-luminescence spectra of some organic dyes22–24, lanthanide complexes29,30 and colloidal nanocrystals5–7,27,31,35,36,39,40 used in LSCs. The combination of a large solar coverage and small spectral overlap between the absorption and photoluminescence spectrum is obtained only in nanocrystals with a care-fully engineered Stokes shift5–7. c | Maximum solar coverage obtainable with

the chromophores in b, calculated in the ‘infinite concentration’ limit. The darker-shaded portion of the bars for the nanocrystal systems indicates the coverage that can be achieved if bulk semiconductors are used. The horizontal dashed lines indicate the spectral coverage of the indicated photo voltaic (PV) cells. Inset: Commission Internationale de l’Éclairage (CIE) coordinates of the portion of solar light transmitted by PV windows embed-ding an LSC panel with the spectral coverage reported in the main panel. For the colloidal nanocrystals, the coordinates refer to the quantum- confined systems. c-Si, crystalline silicon; DCM, 4-dicyanomethyl- 6-dimethylaminostiryl-4H-pyran; TTA, thenoyltrifluoroacetone; TTPO, triphenylphosphine oxide. Panel b is adapted with permission from REF. 5, Macmillan Publishers Limited (CdSe/CdS); REF. 6, Macmillan Publishers Limited (Si); REF. 7, Macmillan Publishers Limited (CuInSeS); REF. 23, The Optical Society (Lumogen Red); REF. 24, The Optical Society (DCM); REF. 27, American Institute of Physics (CdSe and PbS); REF. 29, Royal Society of Chemistry (Oxyminopyrazole-Yb); REF. 30, Elsevier [Eu(TTA)3(TTPO)2]; REF. 31, Wiley-VCH (PbS/CdS); REF. 35, American Chemical Society (Mn:ZnSe); REF. 36, American Chemical Society (Cu:CdSe); and REF. 40, Royal Society of Chemistry (CuInS2).

PERSPECT IVES

NATURE REVIEWS | MATERIALS VOLUME 2 | ARTICLE NUMBER 17072 | 5

Lateral size (cm) 0 20 40 60 80 100

tica

l eci

ency

10

11

12

13

14

15

16

Lateral size (cm)0 20 40 60 80 100

tica

l eci

ency

0

2

4

6

8

10

12

14

umber of reabsor tionevents

umber of reabsor tionevents

0 1 2 3 4 5 6 0 1 2 3 4 5 60

50

100

avelength nm avelength nm500 700 800 1,000 1,200

Abs

orti

on c

oeci

ent

cm–1

), no

rmal

ied

em

issi

onro

babi

lity

0.0

0.5

CuInSeS nanocrystalsLumogen Red

Absor tion

a b

c d

avelength nm400 800 1,200

Abs

orti

on c

oeci

ent

cm–1

)

10–5

10–3

10–4

10–2

10–1

100

101

102

1.0

PMMA luorinated MMA

oda lime glass glass

PMMA luorinated MMA

oda lime glass glass

d = 2 cmd = 25 cmd = 1 m

b d nanocrystalsu n e nanocrystalsi nanocrystalsumogen ed

ntrinsicd = 2 cmd = 25 cmd = 1 m

samples of possible light paths are randomly traced) for LSCs with a thickness of 0.5 cm and identical spectral coverage of 30% of the incident solar power. An ideal photo-luminescence quantum yield of 100% was considered for both systems. For simplicity, in this illustrative comparison,

tail and the high-energy portion of the photoluminescence spectrum. This leads to strong reabsorption in small-area samples (2 cm × 2 cm). Accordingly, the statistical analysis of light propagation indicates that over 50% of the photons collected at the LSC edge experience at least one reabsorption

we neglect the effects of light absorption by the waveguide matrix, which are discussed in more detail below. For Lumogen Red, achieving a 30% spectral coverage requires a high concentration of chromophores in the waveguide, which results in a very large overlap between the low-energy absorption

Figure 3 | Performances achievable with different chromophores and

waveguide materials. a | Absorption spectra of luminescent solar concen-trators (LSCs, device thickness 0.5 mm) with 30% spectral coverage embed-ding Lumogen Red (from REF. 45) or CuInSeS nanocrystals (from REF. 7). The experimental intrinsic emission spectra (black lines) are compared with sim-ulated normalized photoluminescence spectra emitted from the device edges of square LSCs with increasing lateral size d; the spectra were calcu-lated using Monte Carlo ray tracing. The absorption coefficient for Lumogen Red is resonant with the emission maximum at 600 nm and is α = 6.2 cm−1; for CuInSeS nanocrystals, it is resonant with the emission maximum at 965 nm and is α = 0.03 cm−1. The lower panels show the probability for the output photons to have undergone a certain number of absorption and re-emission events. b | Optical efficiency obtained from Monte Carlo simulations for

square LSCs with increasing lateral size embedding Lumogen Red45 or col-loidal nanocrystals of silicon6, CuInSeS (REF. 7) or PbS/CdS (REF. 31). The sim-ulations were performed using the absorption and emission spectra reported in REFS 6,7,31,45. The spectral coverage is set to 30% of the total solar flux for all cases. c | Absorption coefficients of different waveguide materials: conventional and fluorinated poly(methyl methacrylate) (PMMA)49, soda lime glass48 and borosilicate crown (N-BK7) glass50. The shaded area highlights the spectral region most relevant for LSCs using crystalline silicon solar cells. d | Optical efficiency simulated with the Monte Carlo ray tracing method for square LSCs with increasing lateral size embedding silicon nanocrystals in different waveguide materials. The spectral coverage is set to 30% of the total solar power for all cases and the emission wavelength is 830 nm. Panel a is adapted with permission from REF. 45, American Chemical Society.

PERSPECT IVES

6 | ARTICLE NUMBER 17072 | VOLUME 2 www.nature.com/natrevmats

event. Such a strong reabsorption results in a clear cut-off filtering effect in the 25 cm × 25 cm devices, in which as much as 64% of collected photons are re-radiated following reabsorption. Because of the relatively sharp absorption edge, the effect fully saturates in larger LSCs, as indicated by the nearly perfect resemblance between the emission profile and photon statistics of the 1 m × 1 m and 25 cm × 25 cm LSC devices (the fraction of collected photons that undergo at least one re-absorption event is ~68%). The behaviour of LSCs embedding CuInSeS nanocrystals is strikingly different. The small energy gap in the near-infrared region and the type-B emission enables the achievement of the spectral coverage of Lumogen Red with a much smaller spectral overlap between the absorption and emission profiles. As a result, over 95% of collected photons in the 2 cm × 2 cm device are unaffected by reabsorption, and even in the largest LSC, re-radiation accounts for less than 40% of the collected photons. Accordingly, optical distances of at least 25 cm are needed to observe a measurable redshift of the photoluminescence spectrum. However, unlike in Lumogen Red, the residual spectral overlap due to the smooth absorption tail of CuInSeS nanocrystals results in the progressive reabsorption of the propagating luminescence with increasing device size, with no clearcut saturation of the spectral redshift.

The different light-propagation mechanisms in LSCs embedding type-A or type-B chromophores have important consequences on the evolution of the optical efficiency (η) with increasing device dimensions. This is highlighted by a comparison of the simulated optical power efficiency, defined as the ratio between the number of photons extracted from the waveguide edges and the incident photon flux for square LSCs with increasing size and equal spectral coverage embedding Lumogen Red45 or colloidal nanocrystals of CuInSeS (REF. 7), PbS/CdS (REF. 31) or silicon6 (FIG. 3b). In all cases, the photoluminescence efficiency is assumed to be 100% and absorption by the waveguide material is neglected. For Lumogen Red, owing to the strong reabsorption associated with the high-energy shoulder of the emission spectrum, the maximum optical power efficiency for a 2 cm × 2 cm LSC is η = 12.4%. However, as a result of the relatively sharp absorption edge, enabling the reabsorption-free propagation of low-energy luminescence to the device edges, η plateaus to a constant value of ~10.5% for larger devices.

material for the particle shell, which also helps to protect the optical properties from the radical initiators used in the mass polymerization of the nanocomposite polymeric waveguides5–7,37.

The optical efficiencies reported in FIG. 3b show, for all the considered chromophores and device sizes, output powers exceeding 100 W m−2, which can be further increased by lowering the LSC transparency and/or optimizing the device geometry as a function of the employed emitter. Even when considering additional losses in the optical-to-electrical power conversion, this means that with the recently developed chromophores, reabsorption is no longer the main factor hindering a widespread diffusion of the LSC technology. Another important aspect to consider for optimizing the performance of LSCs for BIPV applications is the choice of the waveguide material46,47, which could determine relevant optical losses in large-area devices by absorbing the propagating luminescence (BOX 2).

Optimization of the waveguideThe role of the waveguide material in the optical propagation process has long been overlooked because of the dominant effect of reabsorption by traditional chromophores, which leads to severe optical losses even for short optical distances. The suppression of reabsorption losses in LSCs embedding colloidal nanocrystals with engineered Stokes shift described above has recently enabled the fabrication of LSCs so large that light absorption by the optical waveguide can no longer be neglected6,7,37; the first studies discussing this problem have been published recently46,47. To highlight the differences between various materials for LSC waveguides (FIG. 3c), we compare the absorption coefficient (αWG) of plastic matrices, such as optical poly(methyl methacrylate) (PMMA) and its fluorinated analogue used in low-loss plastic optical fibres, and of inorganic glasses that could be used as substrates in thin-film LSC architectures, namely, soda lime glass (common window glass) and optical-grade borosilicate crown glass (N-BK7). Notably, although all these materials are commonly considered transparent, their absorption cross sections in the 800–1,100 nm spectral region — the region relevant for LSCs equipped with silicon solar cells — differ by orders of magnitude: αWG ≈ 0.5 cm−1 for soda lime glass48 and 3 × 10−2 cm−1 in non-fluorinated PMMA (this value is due to vibrational absorption by the C–H modes, which are absent in the

The dependence of η on the LSC size is markedly different in LSCs embedding type-B chromophores, which show systematically higher initial optical efficiency than Lumogen Red owing to the lower reabsorption at comparable spectral coverage and the continuous gradual decrease in η with increasing device size; there is no saturation owing to the residual weak absorption that is resonant with the photoluminescence spectrum. The same behaviour is observed for silicon nanocrystals.

Strategies to maximize the efficiencyIn light of the above considerations concerning the desired spectral coverage and suppression of the absorption–emission overlap in LSC chromophores, we can now rationalize possible future strategies for achieving highly efficient luminescent solar concentration for BIPV. To maximize the efficiency of LSCs with type-A emitters, it is pivotal to sharpen the absorption edge and increase the Stokes shift to reduce the spectral resonance responsible for short-distance optical losses. For organic dyes, this could be achieved by reducing the vibrational disorder in stiff structures while concomitantly allowing sufficient molecular distortion in the excited state to ensure a large Stokes shift. Because the majority of the output photons are produced by multiple re-emission processes, it is also important that type-A chromophores exhibit an emission efficiency as close to unity as possible. Otherwise, cumulative nonradiative decay losses completely suppress the output probability, even for relatively small LSCs. On the other hand, the optimization of the efficiency of large-area LSCs based on type-B chromophores requires the suppression of the spectral overlap by reducing the bandwidth of both the lowest-energy absorption feature and the emission spectrum, which is mostly due to the polydisperse crystal size. Because of the weaker occurrence of reabsorption and re-radiation events, good device performances can be obtained even with type-B chromophores with non-unity photoluminescence quantum yields, as proven by the high optical power output experimentally observed in CuInSeS or silicon nanocrystals, which featured ~50% photoluminescence efficiency6,7. Nevertheless, optimization of the emission efficiency is an obvious step for maximizing η. In colloidal nanostructures, this is typically achieved by proper molecular passivation and by using a wide-energy-gap

PERSPECT IVES

NATURE REVIEWS | MATERIALS VOLUME 2 | ARTICLE NUMBER 17072 | 7

fluorinated polymer, resulting in a lower absorption coefficient of 2 × 10−4 cm−1)49; N-BK7 glass is also highly transparent, with αWG = 1 × 10−3 cm−1 (REF. 50). Such large differences between the matrix absorption coefficients have profound effects on the solar concentration performance, as revealed by Monte Carlo ray tracing simulations predicting the optical efficiency as a function of device size for square LSCs with a thickness of 0.5 cm based on different matrix materials and embedding silicon nanocrystals (FIG. 3d). This analysis highlights the unsuitability of standard glass as waveguide material for large-area LSCs because its absorption of the propagating luminescence lowers the output probability by over 80% in just a few tens of centimetres. Conventional PMMA has a considerably higher transparency but its (at least partial) fluorination is highly recommended to lower the optical losses in large-area devices. Clearly, fully fluorinated PMMA or N-BK7 glass would yield the best LSC performances. To date, these materials are used in small quantities in top-notch optical technologies but are too expensive for their widespread use in BIPV. However, the growth of a multi-billion dollar market for solar glazing systems (BOX 1) could provide the driving force for a substantial price reduction. The role of the waveguide material in luminescence absorption suggests important methodological guidelines for the correct optical characterization of LSCs. The background absorption spectrum should be measured with extreme care and evaluated in absolute terms. When reporting the absorption spectrum, no arbitrary subtraction of the background should be done, as this neglects the role of the matrix material and of possible extinction contributions due to scattering. Moreover, the absorption spectra of the chromophores, waveguide and LSC device should be rigorously reported in absolute units (cm−1). Normalization of the absorption data or their expression in arbitrary units should be avoided because this hinders the quantitative evaluation of the light-propagation efficiency.

Conclusions and perspectivesIn summary, thanks to recent advancements in the design of colloidal nanostructures, the long-awaited goal of combining suppressed reabsorption with broadband spectral coverage has been achieved. The LSC technology is now set on a clear path to become a PV technology that can be ‘invisibly’ integrated into buildings in the form of colourless and electrodeless solar windows or as other semi-transparent

3. Debije, M. G. Solar energy collectors with tunable transmission. Adv. Funct. Mater. 20, 1498–1502 (2010).

4. van Sark, W. G. J. H. M. Luminescent solar concentrators — a low cost photovoltaics alternative. Renewable Energy 49, 207–210 (2013).

5. Meinardi, F. et al. Large-area luminescent solar concentrators based on Stokes-shift-engineered nanocrystals in a mass-polymerized PMMA matrix. Nat. Photonics 8, 392–399 (2014).

6. Meinardi, F. et al. Highly efficient luminescent solar concentrators based on earth-abundant indirect-bandgap silicon quantum dots. Nat. Photonics 11, 177–185 (2017).

7. Meinardi, F. et al. Highly efficient large-area colourless luminescent solar concentrators using heavy-metal-free colloidal quantum dots. Nat. Nanotechnol. 10, 878–885 (2015).

8. Weber, W. H. & Lambe, J. Luminescent greenhouse collector for solar radiation. Appl. Opt. 15, 2299–2300 (1976).

9. Goetzberger, A. & Greube, W. Solar energy conversion with fluorescent collectors. Appl. Phys. A Mater. Sci. Process. 14, 123–139 (1977).

10. Saifullah, M., Gwak, J. & Yun, J. H. Comprehensive review on material requirements, present status, and future prospects for building-integrated semitransparent photovoltaics (BISTPV). J. Mater. Chem. A 4, 8512–8540 (2016).

11. Debije, M. G. & Verbunt, P. P. C. Solar concentrators: thirty years of luminescent solar concentrator research: solar energy for the built environment. Adv. Energy Mater. 2, 12–35 (2012).

12. Pietryga, J. M. et al. Spectroscopic and device aspects of nanocrystal quantum dots. Chem. Rev. 116, 10513–10622 (2016).

13. Yablonovitch, E. Thermodynamics of the fluorescent planar concentrator. J. Opt. Soc. Am. 70, 1362–1363 (1980).

14. Currie, M. J., Mapel, J. K., Heidel, T. D., Goffri, S. & Baldo, M. A. High-efficiency organic solar concentrators for photovoltaics. Science 321, 226–228 (2008).

15. van Sark, W. G. J. H. M. et al. Luminescent solar concentrators — a review of recent results. Opt. Express 16, 21773–21792 (2008).

16. Banal, J. L., Ghiggino, K. P. & Wong, W. W. H. Efficient light harvesting of a luminescent solar concentrator using excitation energy transfer from an aggregation-induced emitter. Phys. Chem. Chem. Phys. 16, 25358–25363 (2014).

17. Portnoi, M., Sol, C., Tummeltshammer, C. & Papakonstantinou, I. Impact of curvature on the optimal configuration of flexible luminescent solar concentrators. Opt. Lett. 42, 2695–2698 (2017).

18. Aste, N., Tagliabue, L. C., Del Pero, C., Testa, D. & Fusco, R. Performance analysis of a large-area luminescent solar concentrator module. Renewable Energy 76, 330–337 (2015).

19. Slooff, L. H. et al. A luminescent solar concentrator with 7.1% power conversion efficiency. Phys. Status Solidi RRL 2, 257–259 (2008).

20. Debije, M. G., Tzikas, C., Rajkumar, V. A. & de Jong, M. M. The solar noise barrier project: 2. The effect of street art on performance of a large scale luminescent solar concentrator prototype. Renew. Energy 113, 1288–1292 (2017).

21. van Sark, W. et al. The “electric Mondrian” as a luminescent solar concentrator demonstrator case study. Solar RRL 1, 1600015 (2017).

22. Batchelder, J. S., Zewai, A. H. & Cole, T. Luminescent solar concentrators. 1: Theory of operation and techniques for performance evaluation. Appl. Opt. 18, 3090–3110 (1979).

23. Desmet, L., Ras, A. J. M., de Boer, D. K. G. & Debije, M. G. Monocrystalline silicon photovoltaic luminescent solar concentrator with 4.2% power conversion efficiency. Opt. Lett. 37, 3087–3089 (2012).

24. Debije, M. G. et al. Promising fluorescent dye for solar energy conversion based on a perylene perinone. Appl. Opt. 50, 163–169 (2011).

25. Sanguineti, A. et al. High Stokes shift perylene dyes for luminescent solar concentrators. Chem. Commun (Camb.) 49, 1618–1620 (2013).

26. Krumer, Z. et al. Tackling self-absorption in luminescent solar concentrators with type-II colloidal quantum dots. Sol. Energy Mater. Sol. Cells 111, 57–65 (2013).

27. Shcherbatyuk, G. V., Inman, R. H., Wang, C., Winston, R. & Ghosh, S. Viability of using near infrared PbS quantum dots as active materials in luminescent solar concentrators. Appl. Phys. Lett. 96, 191901 (2010).

architectural elements. However, this final integration step and, in particular, the optimization of the power efficiency for devices on the metre scale will require scientists and engineers to face new emerging challenges. On the materials side, the realization of high-quality optical waveguides will be the most demanding task. It will require a perfect dispersion of the nanocrystals inside the waveguide matrix to avoid detrimental light scattering losses by aggregates or structural imperfections. As discussed above, the suppression of the extinction coefficient due to scattering to values lower than 10−3 cm−1 at the luminescence wavelength of the chromophore is necessary for efficient light propagation. Surface functionalization with ligands that are highly compatible with the waveguide material will probably be the key to success, and once properly optimized, it will enable the maximization of the photolu-minescence quantum yield of the nanocrystal and its stability to the fabrication conditions. The adaptation to LSCs of highly transparent polymers or glass variants currently used in photonics and telecommunications will probably also be necessary to produce very large LSCs (that is, with a lateral size of more than one metre), as conventional polymers and, to a larger degree, common glass introduce unsustainable optical losses. Finally, the optical-to-electrical conversion by the PV cells at the LSC edges will have to be suitably designed and optimized. This last step in the luminescent solar concentration process has not yet been fully investigated because of the challenges in the realization of highly performing optical waveguides. Together with the role of the waveguide absorption, this will soon be at the forefront of research into LSCs for real-world applications.

Francesco Meinardi and Sergio Brovelli are at the Dipartimento di Scienza dei Materiali, Università degli

Studi di Milano-Bicocca, via Roberto Cozzi 55, I-20125 Milano, Italy; and the Glass to Power Srl, Francesco

Daverio, 6, 20135 Milano, Italy.

Francesco Bruni is at the Glass to Power Srl, Francesco Daverio, 6, 20135 Milano, Italy.

Correspondence to F.M. and S.B. francesco.meinardi@unimib.it;

sergio.brovelli@unimib.it

doi:10.1038/natrevmats.2017.72 Published online 21 Nov 2017

1. Jelle, B. P., Breivik, C. & Røkenes, H. D. Building integrated photovoltaic products: a state-of-the-art review and future research opportunities. Sol. Energ. Mater. Sol. Cells 100, 69–96 (2012).

2. The European Parliament and the Council of the European Union. Directive 2010/31/EU of the European Parliament and the Council on the energy performance of buildings (recast). Off. J. Eur. Union 18, 13–35 (2010).

PERSPECT IVES

8 | ARTICLE NUMBER 17072 | VOLUME 2 www.nature.com/natrevmats

28. Bomm, J. et al. Fabrication and full characterization of state-of-the-art quantum dot luminescent solar concentrators. Sol. Energy Mater. Sol. Cells 95, 2087–2094 (2011).

29. Sanguineti, A. et al. NIR emitting ytterbium chelates for colourless luminescent solar concentrators. Phys. Chem. Chem. Phys. 14, 6452–6455 (2012).

30. Wang, T. et al. Luminescent solar concentrator employing rare earth complex with zero self-absorption loss. Sol. Energy 85, 2571–2579 (2011).

31. Zhou, Y. et al. Near infrared, highly efficient luminescent solar concentrators. Adv. Energy Mater. 6, 1501913 (2016).

32. Bronstein, N. D. et al. Luminescent solar concentration with semiconductor nanorods and transfer-printed micro-silicon solar cells. ACS Nano 8, 44–53 (2013).

33. Talapin, D. V. et al. Seeded growth of highly luminescent CdSe/CdS nanoheterostructures with rod and tetrapod morphologies. Nano Lett. 7, 2951–2959 (2007).

34. Bronstein, N. D. et al. Quantum dot luminescent concentrator cavity exhibiting 30-fold concentration. ACS Photonics 2, 1576–1583 (2015).

35. Erickson, C. S. et al. Zero-reabsorption doped-nanocrystal luminescent solar concentrators. ACS Nano 8, 3461–3467 (2014).

36. Bradshaw, L. R., Knowles, K. E., McDowall, S. & Gamelin, D. R. Nanocrystals for luminescent solar concentrators. Nano Lett. 15, 1315–1323 (2015).

37. Meinardi, F. et al. Doped halide perovskite nanocrystals for reabsorption-free luminescent solar concentrators. ACS Energy Lett. 2, 2368–2377 (2017).

38. Viswanatha, R., Brovelli, S., Pandey, A., Crooker, S. A. & Klimov, V. I. Copper-doped inverted core/shell nanocrystals with “permanent” optically active holes. Nano Lett. 11, 4753–4758 (2011).

39. Hu, X. et al. Ray-trace simulation of CuInS(Se)2 quantum dot based luminescent solar concentrators. Opt. Express 23, A858–A867 (2015).

40. Knowles, K. E., Kilburn, T. B., Alzate, D. G., McDowall, S. & Gamelin, D. Bright CuInS2/CdS

52. Verbunt, P. P. C. et al. Controlling light emission in luminescent solar concentrators through use of dye molecules aligned in a planar manner by liquid crystals. Adv. Funct. Mater. 19, 2714–2719 (2009).

53. Verbunt, P. P. C., Sánchez-Somolinos, C., Broer, D. J. & Debije, M. G. Anisotropic light emissions in luminescent solar concentrators–isotropic systems. Opt. Express 21, A485–A493 (2013).

Author contributionsF.M. researched data for article, provided substantial contri-bution to discussion of the content and reviewed and edited the manuscript before submission. F.B. researched data for article. S.B. researched data for article, provided substantial contribution to discussion of the content and wrote, reviewed and edited the manuscript before submission.

Competing interests statementThe authors declare no competing interests.

Publisher’s noteSpringer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

How to cite this articleMeinardi, F., Bruni, F. & Brovelli, S. Luminescent solar con-centrators for building-integrated photovoltaics. Nat. Rev. Mater. 2, 17072 (2017).

nanocrystal phosphors for high-gain full-spectrum luminescent solar concentrators. Chem. Commun. (Camb.) 51, 9129–9132 (2015).

41. Chen, W. et al. Heavy metal free nanocrystals with near infrared emission applying in luminescent solar concentrator. Solar RRL 1, 1700041 (2017).

42. Meier, C., Gondorf, A., Lüttjohann, S., Lorke, A. & Wiggers, H. Silicon nanoparticles: absorption, emission, and the nature of the electronic bandgap. J. Appl. Phys. 101, 103112 (2007).

43. Hannah, D. C. et al. On the origin of photoluminescence in silicon nanocrystals: pressure-dependent structural and optical studies. Nano Lett. 12, 4200–4205 (2012).

44. Zhao, Y. & Lunt, R. R. Transparent luminescent solar concentrators for large-area solar windows enabled by massive Stokes-shift nanocluster phosphors. Adv. Energy Mater. 3, 1143–1148 (2013).

45. Colby, K. A. et al. Electronic energy migration on different time scales: concentration dependence of the time-resolved anisotropy and fluorescence quenching of Lumogen Red in poly(methyl methacrylate). J. Phys. Chem. A 114, 3471–3482 (2010).

46. Zettl, M., Mayer, O., Klampaftis, E. & Richards, B. S. Investigation of host polymers for luminescent solar concentrators. Energy Technol. 5, 1037–1044 (2017).

47. Kastelijn, M. J., Bastiaansen, C. W. M. & Debije, M. G. Influence of waveguide material on light emission in luminescent solar concentrators. Opt. Mater. 31, 1720–1722 (2009).

48. Rubin, M. Optical properties of soda lime silica glasses. Sol. Energy Mater. 12, 275–288 (1985).

49. Ioannides, N. et al. Approaches to mitigate polymer-core loss in plastic optical fibers: a review. Mater. Res. Express 1, 032002 (2014).

50. SCHOTT. Optical glass data sheets SCHOTT https://refractiveindex.info/download/data/2015/schott-optical-glass-collection-datasheets-july-2015-us.pdf (2015).

51. Batchelder, J. S., Zewail, A. H. & Cole, T. Luminescent solar concentrators. 2: Experimental and theoretical analysis of their possible efficiencies. Appl. Opt. 20, 3733–3754 (1981).

FURTHER INFORMATIONEnergy Technology Plan: http://ec.europa.eu/energy/en/publications/towards-integrated-strategic-energy-technology-set-plan-accelerating-european-energyGlobal Industry Analysts, Inc. report: http://www.strategyr.com/Building_Integrated_Photovoltaics_BIPV_Market_Report.asp#sthash.fE17VTUo.dpbsn-tech Research report: http://ntechresearch.com/market_reports/bipv-technologies-and-markets-2017-2024

ALL LINKS ARE ACTIVE IN THE PDF

PERSPECT IVES

NATURE REVIEWS | MATERIALS VOLUME 2 | ARTICLE NUMBER 17072 | 9