Transparent photovoltaic cells

Abstract

A transparent photovoltaic cell and method of making are disclosed. The photovoltaic cell may include a transparent substrate and a first active material overlying the substrate. The first active material may have a first absorption peak at a wavelength greater than about 650 nanometers. A second active material is disposed overlying the substrate, the second active material having a second absorption peak at a wavelength outside of the visible light spectrum. The photovoltaic cell may also include a transparent cathode and a transparent anode.

Claims

1. A transparent tandem cell photovoltaic device comprising: a transparent substrate; a transparent anode; a first transparent photovoltaic cell including: a first transparent active layer overlying the transparent substrate, the first transparent active layer having an absorption peak at a wavelength greater than 650 nanometers, the absorption peak of the first transparent active layer being greater than an absorption of the first transparent active layer at any wavelength between about 450 and 650 nanometers; and a second transparent active layer overlying the transparent substrate, the second transparent active layer having an absorption peak at a wavelength between 300 and 450 nanometers or between 650 and 2500 nanometers, the absorption peak of the second transparent active layer being greater than an absorption of the second transparent active layer at any wavelength between about 450 and 650 nanometers; a second transparent photovoltaic cell including: a third transparent active layer overlying the transparent substrate, the third transparent active layer having an absorption peak at a wavelength greater than 650 nanometers, the absorption peak of the third transparent active layer being greater than an absorption of the third transparent active layer at any wavelength between about 450 and 650 nanometers; and a fourth transparent active layer overlying the transparent substrate, the fourth transparent active layer having an absorption peak at a wavelength between 300 and 450 nanometers or between 650 and 2500 nanometers, the absorption peak of the fourth transparent active layer being greater than an absorption of the fourth transparent active layer at any wavelength between about 450 and 650 nanometers; a transparent recombination zone disposed between the first transparent photovoltaic cell and the second transparent photovoltaic cell; and a transparent cathode; the transparent tandem cell photovoltaic device having at least one absorption peak at a wavelength greater than about 650 nanometers that is greater than an absorption of the transparent tandem cell photovoltaic device at any wavelength between about 450 and 650 nanometers, wherein the transparent tandem cell photovoltaic device is operable to transmit visible light with wavelengths between 4501 and 650 nanometers.

2. The transparent tandem cell photovoltaic device of claim 1 wherein the transparent tandem cell photovoltaic device is characterized by an absorption peak at a wavelength less than 450 nm that is greater than the absorption of the transparent tandem cell photovoltaic device at any wavelength between about 450 and 650 nanometers.

3. The transparent tandem cell photovoltaic device of claim 1 wherein: the first transparent active layer comprises at least one of a phthalocyanine, a porphyrin, a naphthalocyanine, or nanotubes; and the third transparent active layer comprises at least one of a phthalocyanine, a porphyrin, a naphthalocyanine, or nanotubes.

4. The transparent tandem cell photovoltaic device of claim 1 wherein: the second transparent active layer comprises at least one of carbon 60 (C.sub.60) or a nanotube; and wherein the fourth transparent active layer comprises at least one of carbon 60 (C.sub.60) or a nanotube.

5. The transparent tandem cell photovoltaic device of claim 1 wherein the transparent recombination zone comprises at least one of indium tin oxide (ITO), bathocuproine (BCP) or molybdenum oxide (MoOx).

6. The transparent tandem cell photovoltaic device of claim 1 further comprising: a third transparent photovoltaic cell including: a fifth transparent active layer overlying the transparent substrate, the fifth transparent active layer having an absorption peak at a wavelength greater than 650 nanometers, the absorption peak of the fifth transparent active layer being greater than the absorption of the fifth transparent active layer at any wavelength between about 450 and 650 nanometers; and a sixth transparent active layer overlying the transparent substrate, the sixth transparent active layer having an absorption peak at a wavelength between 300 and 450 nanometers or between 650 and 2500 nanometers, the absorption peak of the sixth transparent active layer being greater than the absorption of the sixth transparent active layer at any wavelength between about 450 and 650 nanometers; and a second transparent recombination zone disposed between the second transparent photovoltaic cell and the third transparent photovoltaic cell.

7. The transparent tandem cell photovoltaic device of claim 1 further comprising a transparent near-infrared (NIR) mirror.

8. The transparent tandem cell photovoltaic device of claim 1 wherein the transparent anode comprises at least one of: Indium Tin Oxide (ITO), molybdenum oxide (MoOx), fluorinated tin oxide, poly(3,4-ethylenedioxythiophene), Poly (3,4-ethylenedioxythiophene)-poly(styrenesulfonate), gallium doped zinc oxide, or aluminum doped zinc oxide.

9. The transparent tandem cell photovoltaic device of claim 1 wherein transparent comprises an average visible transparency of equal to or greater than 45%.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1A shows a schematic of a control solar cell;

(2) FIG. 1B shows a schematic of a full transparent solar cell embodiment;

(3) FIG. 1C is a graph showing the extinction coefficient, k, of the active layers shown in FIGS. 1A and 1B;

(4) FIG. 1D is a graph showing the current-voltage (J-V) curves for the ClAlP.sub.c-C.sub.60 control and transparent cells shown in FIGS. 1A and 1B;

(5) FIG. 2A is a graph showing photocurrent increase by a factor of 3 at an optimum thickness of 120 nm so that .sub.p increases by nearly the same amount;

(6) FIG. 2B is a graph showing the series resistance diminish and the fill factor (FF) saturates close to the value for the control cell as Indium Tin Oxide (ITO) thickness is increased;

(7) FIG. 3A is a graph showing external quantum efficiency (EQE) as a function of wavelength for several thicknesses of ITO and control layers;

(8) FIG. 3B is a graph showing transmission % as a function of wavelength for several thicknesses of ITO and control layers;

(9) FIG. 3C shows the measured solar simulator spectrum exhibiting characteristics of the Xe-lamp and NREL reported mc-Si external quantum efficiency (EQE) for the reference-diode used to measure the solar simulator intensity;

(10) FIG. 3D shows the measured and calculated reflectivity of the distributed Bragg reflector used in this study as the transparent, NIR mirror;

(11) FIGS. 4A and 4B show solar cell arrays positioned in front of a picture of a rose to highlight the transparency of the fully assembled device;

(12) FIG. 4C shows a solar cell array coupled to an LCD clock;

(13) FIGS. 4D and 4E show an alternate embodiment of a solar cell array positioned in front of a picture of a mountain to highlight the transparency of the fully assembled device;

(14) FIG. 4F is a picture of a full circuit assembly with connections to an LCD clock.

(15) FIG. 5A is a graph showing external quantum efficiency (EQE) as a function of wavelength for a SnPc device;

(16) FIG. 5B is a graph showing transmission % as a function of wavelength for a SnPc device;

(17) FIG. 6A is a graph showing a comparison between SnPc and ClAlPc designs;

(18) FIG. 6B is a graph showing the effect of ITO cathode thickness;

(19) FIGS. 6C and 6D show the transfer matrix simulations of the average visible transmission (AVT, left column) and short-circuit current (right column) of the transparent OPV architecture as a function of the anode and cathode ITO thicknesses without a NIR mirror;

(20) FIGS. 6E and 6F show the transfer matrix simulations of the average visible transmission (AVT, left column) and short-circuit current (right column) of the transparent OPV architecture as a function of the anode and cathode ITO thicknesses with a NIR mirror;

(21) FIG. 7 is a block diagram of a device with a mixed layer including both a donor and an acceptor;

(22) FIG. 8 is a block diagram of a tandem device;

(23) FIGS. 9A and 9B are graphs showing different bandgaps that may be used to optimize a tandem device;

(24) FIGS. 10A and 10B are graphs showing practical efficiency limits of several of the embodiments disclosed herein;

(25) FIG. 11 is a diagram showing the solar flux and the photopic response of the human eye; and

(26) FIG. 12 is a diagram showing an e-reader, smart phone and display screen including a photovoltaic array as disclosed herein.

DETAILED DESCRIPTION OF THE INVENTION

(27) Described herein are improved transparent solar cell designs, e.g., transparent organic photovoltaic devices (TOPV). The term transparent as used herein encompasses an average visible transparency of a straight through beam of 45% or more. The term semi-transparent as used herein encompasses an average visible transparency of a straight through beam of approximately 10%-45%. In general, the designs include molecular active layers with strong absorption features outside of the visible light spectrum, e.g., in the ultra-violet (UV) and/or near-infrared (NIR) solar spectrum. The devices may include selective high-reflectivity NIR and broadband anti-reflection contact coatings. Devices may be formed as heterojunction solar cells with an organic active layer, such as chloroaluminum phthalocyanine (ClAlPc) or SnPc as a donor and a molecular active layer such as C.sub.60 acting as an acceptor and having peak-absorption in the UV and NIR solar spectrum. Other suitable materials for the active layers include any suitable phthalocyanine, porphyrin, naphthalocyanine dye, carbon nanotubes or molecular excitonic materials with absorption peaks outside the visible spectrum. Such devices may be formed in a tandem structure with one or more subcells joined via a recombination zone. Such devices may be used in a variety of applications including rigid and flexible computer display screens used in a desktop monitor, laptop or notebook computer, tablet computer, mobile phone, e-readers and the like. Other applications include watch crystals, automotive and architectural glass including sunroofs and privacy glass. The photovoltaic devices may be used for active power generation, e.g., for completely self-powered applications, and battery charging (or battery life extension).

(28) Near-infrared (NIR) as recited herein is defined as light having wavelengths in the range from about 650 to about 850 nanometers (nm). Ultraviolet (UV) as recited herein is defined as light having wavelengths less than about 450 nm. The use of an active layer having absorption in the NIR and the UV allows for the use of selective high-reflectivity near-infrared mirror coatings to optimize device performance while also permitting high transmission of visible light through the entire device. Visible light as recited herein is defined as light having wavelengths to which the human eye has a significant response, from about 450 to about 650 nm.

(29) In one embodiment, devices were fabricated on 150 nm of patterned Indium Tin Oxide (ITO) (15 /sq.) pre-coated onto glass substrates. The ITO is one component of an electrode. The ITO was solvent-cleaned and subsequently treated in oxygen plasma for 30 seconds immediately prior to loading into a high vacuum chamber (<110.sup.6 Torr). ClAlPc and C.sub.60 were purified once by vacuum train sublimation prior to loading. Bathocuproine (BCP) and molybdenum trioxide (MoO.sub.3) were used as purchased. MoO.sub.3 is another component of an electrode. The MoO.sub.3 (20 nm), ClAlPc (15 nm), C.sub.60 (30 nm), BCP (7.5 nm), and a 100 nm thick Ag cathode were sequentially deposited via thermal evaporation at a rate of 0.1 nm/s. The top ITO cathode for the transparent devices was rf-sputtered directly onto the organic layers at low power (7-25 W) with 10 sccm Ar flow (6 mTorr) and 0.005-0.03 nm/second. Cathodes were evaporated through a shadow mask, defining a 1 millimeter (mm)1.2 mm active device area. A near-infrared distributed Bragg reflector (DBR) utilized as the transparent NIR mirror was grown separately on quartz via sputtering of 7 alternating layers of TiO.sub.2 and SiO.sub.2 at about 0.1 nm/second with thicknesses centered around a wavelength of 800 nm (200 nm stop band). Broad-band antireflection (BBAR) coatings precoated on quartz substrates (1-side) were attached to the DBRs via index matching fluid to reduce additional glass/air interface reflections. Transmission data of the assembled devices were obtained at normal incidence with a Cary Eclipse 5000 dual-beam spectrophotometer without reference samples. Current density versus voltage (J-V) characteristics were measured in the dark and under simulated AM1.5G solar illumination without solar mismatch correction (for reference, the mismatch factor was estimated to be 1.05) and external quantum efficiency (EQE) measurements were collected utilizing an NREL calibrated Si detector. Optical interference modeling was carried out according to the method of L. A. A. Pettersson, L. S. Roman, and O. Inganas, Journal of Applied Physics 86, 487 (1999), the contents of which are herein incorporated by reference. The exciton diffusion lengths of ClAlPc and C.sub.60 were estimated from fitting the magnitudes of the photocurrent and EQE to be 53 nm and 105 nm, respectively.

(30) FIG. 1A shows a schematic of a control solar cell 10. The control solar cell includes a substrate 11, an anode 12, a donor layer 13 e.g., ClAlPc, a molecular active layer, e.g., C.sub.60, acting as an acceptor layer 14 and a cathode 15. In this example, the anode 15 is opaque, e.g., silver. FIG. 1B shows a schematic of a full transparent solar cell 20. The device 20 generally includes a transparent substrate 21, an anode 22, a donor layer 23, e.g., ClAlPc, a molecular active layer, e.g., C.sub.60, acting as an acceptor layer 24, and a cathode 25. The donor layer 23 and the acceptor layer 24 have absorption peaks in the ultra-violet (UV) and near-infrared (NIR) spectrum. In this example, the substrate is quartz. It should be understood that a variety of rigid and flexible substrates may be used. For example, the substrate may be glass, a rigid or flexible polymer, e.g., a screen protector or skin, or may be combined with other layers such as encapsulating layers, anti-reflecting layers or the like. In this example, the transparent anode 22 and cathode 25 are formed of conducting oxide, e.g. ITO/MoO.sub.3. It should be understood that the anode 22 and cathode 25 may be formed of other materials such as tin oxides, fluorinated tin oxides, nanotubes, Poly(3,4-ethylenedioxythiophene) (PDOT) or PEDOT:PSS (Poly(3,4-ethylenedioxythiophene)poly(styrenesulfonate)), gallium doped zinc oxide, aluminum doped zinc oxide and other materials having suitable transparency and conductivity. The device 20 may also include a near-infrared DBR 26 and one or more broad-band antireflection (BBAR) coatings 27.

(31) FIG. 1C is a graph showing the extinction coefficient, k, of the active layers shown in FIGS. 1A and 1B. FIG. 1D is a graph showing the current-voltage (J-V) curves for the ClAlPc-C.sub.60 control and transparent cells of FIGS. 1A and 1B for a range of thicknesses of ITO. The absorption peak for ClAlPc is positioned in the NIR range (740 nm). This allows for the incorporation of a NIR reflecting mirror and simultaneous optimization of the solar cell performance and visible-transmissivity as diagramed in FIGS. 1A and 1B. It should be understood that the donor and/or acceptor layers may have one or more absorption peaks outside of the visual spectrum. In this example, the ClAlPc also has a second absorption peak in the UV range. A summary of various device performances is provided in Table 1.

(32) TABLE-US-00001 TABLE 1 Cathode Thickness Cathode Jsc Voc FF .sub.p AVT (nm) Composition (mA/cm.sup.2) (V) (%) (%) 100 Ag 4.7 0.77 0.55 2.4 0 20 ITO 1.5 0.69 0.39 0.5 67 120 ITO 3.2 0.71 0.46 1.3 65 20 ITO/NIR mir. 2.2 0.73 0.32 0.6 53 40 ITO/NIR mir. 2.5 0.71 0.49 1.1 55 80 ITO/NIR mir. 2.9 0.71 0.46 1.2 56 120 ITO/NIR mir. 4.4 0.71 0.44 1.7 56 170 ITO/NIR mir. 3.2 0.69 0.48 1.3 66

(33) Table 1 generally includes data showing the performance of control OPVs with an Ag cathode, transparent OPVs with ITO cathode, and OPVs with ITO cathode and NIR mirror, at 0.8 sun illumination corrected for solar spectrum mismatch. Short circuit current, JSC, open circuit voltage, VOC, fill factor, FF, power conversion efficiency, .sub.p, and the average visible transmission, AVT, are indicated. The control device with a thick Ag cathode exhibits a power conversion efficiency (.sub.p) of 1.90.2%, open circuit voltage (Voc)=0.800.02V, short-circuit current density (Jsc)=4.70.3 mA/cm.sup.2, and fill-factor (FF)=0.550.03, which is comparable to previous reports.

(34) When the Ag cathode of the control cell is replaced with ITO, the short-circuit current Jsc drops significantly to 1.50.1 mA/cm..sup.2, the FF drops to 0.350.02, and the open-circuit voltage Voc decreases slightly to 0.70.02V leading to .sub.p=0.40.1%. The FF decreases due to an increase in series resistance from the thin ITO that is observable in the J-V curve under forward bias in FIG. 1C. FIG. 2B is a graph showing the series resistance diminish and the FF saturate close to the value for the control cell as ITO thickness is increased. In FIGS. 2A and 2B, the solid lines are from actual simulations, the dashed lines are simply guides to the eye. The slight drop in Voc, independent of ITO thickness, is likely due to a slight reduction in the cathode-anode work function offset. Nonetheless it is remarkable that when utilizing ITO as both anode and cathode there is enough deposition anisotropy in the work function to support this large Voc and is likely assisted by the large work function MoO.sub.3 layer.

(35) The Jsc decreases as the cathode is switched from Ag to ITO due to reduced cathode reflections that reduce the total absorption across the spectrum in the active layers. FIG. 2A is a graph showing photocurrent increase by a factor of 3 at an optimum thickness of 120 nm so that .sub.p increases by nearly the same amount. Fitting this data with the optical interference model shows that this behavior stems from interference of the backside ITO cathode reflection. FIG. 3A is a graph showing EQE as a function of wavelength for several thicknesses of ITO and control layers with and without NIR reflecting mirrors. The approximate visible photopic range is highlighted by vertical dashed lines. FIG. 3B is a graph showing transmission % as a function of wavelength for several thicknesses of ITO and control layers. Comparing EQE and transmission of the ITO-only devices, the absorption for the thinnest and optimized thicknesses appears equivalent. Inspection of the simulations shows, however, that the NIR field distribution is shifted from within the ITO anode to the ClAlPc active layer as the ITO cathode thickness increases, so that the total transmission appears the same even though the active layer absorption changes substantially. This highlights an important aspect of transparent OPV architectures; despite the seemingly simple optical configuration, interference management is still crucial to device optimization, particularly for NIR absorbing cells and for materials with low exciton diffusion lengths.

(36) Despite the significant impact on the photocurrent, the average visible transmissivity (AVT) shows little variation with ITO thickness (see e.g., FIG. 2A). The optical model predicts a slight decrease in AVT with ITO thicknesses that is not observed experimentally possibly due to model parameter uncertainties or varying optical constants during thicker ITO growths. Optimized cells without the NIR mirror show min (max) transmission values of 50% (74%) at 450 nm (540 nm) and an AVT of 65% (standard deviation of 7%). These transmission values decrease slightly with the incorporation of the NIR reflector to min (max) transmission values of 47% (68%) at 450 nm (560 nm) and an AVT of 56% (standard deviation of 5%), where this reduction results from increased off-resonance visible reflections of the mirror. It is possible to remove the off-resonance reflection oscillations in the visible spectra by designing more complex hot-mirror architectures to improve the AVT closer to that of the cell without the NIR mirror, but this typically requires a greater number layers. Hot mirror architectures are described in A. Thelen, Thin Films for Optical Systems 1782, 2 (1993), which is hereby incorporated by reference. High reflectivity of 99% between 695-910 nm also makes these devices useful for simultaneous NIR rejection in architectural cooling. Additionally, the use of the BBAR coatings next to the DBR (outcoupling) and below the substrates (incoupling), results in a concomitant increase in the quantum efficiency by 2-3% and the AVT by 4-6%.

(37) FIG. 3C shows the measured solar simulator spectrum (left axis) exhibiting characteristics of the Xe-lamp and NREL reported mc-Si external quantum efficiency (EQE) for the reference-diode used to measure the solar simulator intensity (right-axis). Because the responsivity of the reference diode extends significantly beyond the response of the OPV cell, the extra NIR light from the solar simulator (compared to the AM1.5G spectrum) results in solar mismatch factors less than 1. FIG. 3D shows the measured (left axis, circles) and calculated (left axis, solid line) reflectivity of the distributed Bragg reflector used in this study as the transparent, NIR mirror. Also shown is the transmission spectrum (right-axis) of the broad-band antireflection (BBAR) coatings.

(38) To highlight the transparency of the fully assembled device, FIGS. 4A and 4B show solar cell arrays in front of a picture of a rose. Both picture-detail and color-clarity are minimally disrupted so that details of the device array pattern are even difficult to discern. In this example the array has a common cathode 25a and a plurality of anodes 22a. The device also includes an active area 30 which includes the donor layer(s), acceptor layer(s) and reflective mirrors. In this particular example, an array of 10 individual OPV devices is formed on the substrate 21a. FIG. 4C shows the array wired to power an LCD clock. FIGS. 4D and 4E show an alternate embodiment of a solar cell array positioned in front of a picture of a mountain to highlight the transparency of the fully assembled device.

(39) FIG. 4F is a picture of a full circuit assembly (left). Electrical connections are made to the ITO contacts of the OPV device (array) via carbon-tape. The LCD clock is connected to circuitry (right) that limits the voltage and passes excess current to a small LED such that the clock works under a wide range of OPV illumination conditions. The LCD clock requires approximately 1.5V and 10 A and can be run by the solar cell for intensities 0.05 suns (note that under the ambient lighting <0.01 sun, the clock is off).

(40) Optimizing the transparent OPV structure with just the cathode thickness, power conversion efficiency of 1.00.1% is obtained, with a simultaneous average transmission of 663%. Incorporation of the NIR reflector and BBAR coatings with the optimized ITO thickness (see FIG. 2A) improves the power conversion efficiency to 1.40.1% with an average transmission of 562%. With the NIR mirror, the increase in power conversion efficiency stems from additional NIR photocurrent in the ClAlPc layer where the EQE shows a near doubling of the peak ClAlPc EQE from 10% to 18% (see FIG. 3A). The optimized power efficiency is nearly triple that of an existing visible-absorbing, semi-transparent, copper phthalocyanine planar device while also exhibiting 30% more average transmission, but is slightly less efficient (0.75) than semi-transparent bulk-heterojunction structures that gain efficiency from active layer absorption in the visible and subsequently have nearly half the transmission.

(41) Switching from planar to bulk-heterojunctions in these structures, efficiencies of 2-3% may be possible for this material set with nearly identical visible transmission, and is currently under investigation. Tandem stacking of subcells with active layer absorption deeper into the infrared could also enhance these efficiencies; combined with more sophisticated NIR mirrors, efficiencies beyond several percent and average visible transmission >70% are possible.

(42) In another embodiment, SnPc, e.g., SnPc-C.sub.60, may be used to construct transparent solar cells. Solar cell designs based on SnPc may achieve >2% efficient solar cell with >70% transmission of visible light (70% average transmission across visible spectrum). The following layers were used in this example: ITO/SnPc(10 nm)/C60(30 nm)/BCP(10 nm)/ITO(10 nm)/DBR. In this example, the ITO was sputtered directly. The distributed Bragg reflectors (DBR) were applied with index matching fluid (IMF). FIG. 5A is a graph showing the EQE as a function of wavelength for the SnPc device. FIG. 5B is a graph showing transmissivity as a function of wavelength for the full TOPV SnPc device. A summary of various device performances is provided in Table 2:

(43) TABLE-US-00002 TABLE 2 Cathode Jsc Voc FF (%) Ag 6.15 0.40 0.55 1.3 ITO 1.54 0.33 0.48 0.2 ITO-DBR 2.25 0.34 0.44 0.3

(44) The device may include a NIR mirror (transparent to visible light) composed of either metal/oxide (e.g. TiO.sub.2/Ag/TiO.sub.2) or dielectric stacks (DBRs e.g. consisting of SiO.sub.2/TiO.sub.2). Anti-reflection coatings may be composed of single or multilayer dielectric materials. As noted above, the molecular active layer may also be composed of any suitable phthalocyanine, porphyrin, naphthalocyanine dye, carbon nanotube, or molecular excitonic materials with absorption peaks outside of the visible spectrum.

(45) FIG. 6A is a graph showing a comparison between SnPc and ClAlPc reference (opaque) designs. A summary of various device performances is provided in Table 3:

(46) TABLE-US-00003 TABLE 3 Donor Thick Jsc Voc FF (%) SnPc 100 6.15 0.40 0.50 1.2 ClAlPc 200 4.70 0.77 0.55 2.0

(47) FIG. 6B is a graph showing the electric field and the effect of ITO cathode thickness. Calculated optical field, |E|.sup.2, of the transparent OVP as a function of position at a fixed wavelength close to the peak absorption of the ClAlPc active layer (740 nm) for an ITO cathode thickness of 20 nm (black line) and 120 nm (red line). Note the enhancement of the field within the ClAlPc layer for the optimized ITO thickness, where the absorption is proportional to |E|.sup.2 integrated over position. In general, there is a strong dependence on ITO thickness.

(48) FIGS. 6C and 6D show the transfer matrix simulations of the average visible transmission (AVT, left column) and short-circuit current (right column) of the transparent OPV architecture as a function of the anode and cathode ITO thicknesses without a NIR mirror. FIGS. 6(e) and 6(f) show the transfer matrix simulations of the average visible transmission (AVT, left column) and short-circuit current (right column) of the transparent OPV architecture as a function of the anode and cathode ITO thicknesses with a NIR mirror. The vertical dashed line indicates the thickness of the ITO anode utilized in this study. The active layer structure was Anode/MoO.sub.3(20 nm)/ClAlPc(15 nm)/C.sub.60(30 nm)/BCP(7.5 nm)/Cathode where the exciton diffusion lengths of ClAlPc and C.sub.60 were estimated from fitting the magnitudes of the photocurrent and EQE of the control cell to be 84 nm and 156 nm, respectively.

(49) The structure shown in FIG. 1B includes discrete layers for the donor, e.g., ClAlPc or SnPc, and the acceptor, e.g., Cho. It should be understood that the donor and acceptor may be combined in a single or mixed layer as shown generally in FIG. 7. In this embodiment the device 40 may have a mixed layer 46 including both a donor and an acceptor. The mixed layer generally has a thickness d.sub.mixed as shown. The device 40 may optionally include a discrete donor layer 48 and/or acceptor layer 46. The donor layer 48, if present, has a thickness d.sub.Donor as shown. The acceptor 46 layer, if present, has a thickness d.sub.Acceptor as shown. It should be understood that FIG. 7 is simplified for matters of clarity and may include additional layers that are not shown. In this example, the device 40 also includes a transparent cathode 42 and a transparent anode 50. The thicknesses of each layer may be selected as generally outlined above. It should be understood that such a structure may also include other layers including anti-reflective layers and mirror layers as disclosed in the various embodiments herein.

(50) An optimization process may generally be performed as follows:

(51) i) Optimize for d.sub.Donor, d.sub.Acceptor (total);

(52) ii) Fix d.sub.Donor, d.sub.Acceptor (total);

(53) iii) Vary d.sub.mixed;

(54) iv) d.sub.Donor=d.sub.Donor (total)(d.sub.mixed/2);

(55) v) d.sub.Acceptor=d.sub.Acceptor (total)(d.sub.mixed/2); and

(56) vi) Optimize for ratio (d.sub.Donor:d.sub.Acceptor).

(57) For devices having a mixed layer only, optimization may include an adjustment of the thickness of the mixed layer (step iii) and an adjustment of the ratio d.sub.Donor:d.sub.Acceptor (step vi).

(58) FIG. 8 is a block diagram of a tandem device 60. The device 60 generally includes at least a first and second cell 66, 68. Each cell may have the structure generally disclosed above. Each of the first and second cells 66, 68 function as transparent subcells. Each may have a varying NIR spectral responsivity. Each of the first and second cells may have absorption peaks at wavelengths outside of the visible light spectrum. A recombination zone 72a is disposed between the first and second cells 66, 68. The recombination zone may be composed of a variety of compounds including, e.g., ITO(0.5-10 nm), or BCP/Ag(0.1-2 nm)/MoOx. Additional recombination zones are disposed between subsequent pairs of subcells as generally shown by reference number 72b. It should be understood that FIG. 8 is simplified for matters of clarity and may include additional layers that are not shown. In this example, the device 60 also includes a cathode 64 and an anode 70. The device may optionally include a transparent NIR mirror 62. FIGS. 9(a) and 9(b) are graphs showing different bandgaps associated with materials that may be used to optimize a device, e.g., US J. Aggregate (FIG. 9(a)) and carbon nanotubes (FIG. 9(b)).

(59) It should be understood that multiple bandgaps may be selected for successive layers stacked in a tandem device in order to yield a device with the desired efficiency. In such devices, overall transparency is improved over devices that are independently fabricated and post integrated or macroscopically combined. This is possible because such a device benefits from a closely matched index of refraction at each interface between successive layers. The stacked structure may be transparent or semi-transparent.

(60) FIGS. 10A and 10B are graphs showing practical efficiency limits of several of the embodiments disclosed herein. FIG. 11 is a diagram showing solar flux and the photopic response of the human eye. In general, the photopic response of the human eye peaks in the green spectrum 530-500 nm and tapers off below 450 nm and above 650 nm.

(61) FIG. 12 is a diagram showing an e-reader 80, smart phone 82 and display screen 84 including photovoltaic arrays 86, 88 and 90 disposed on their respective display screens. It should be understood that a variety of devices may incorporate the photovoltaic devices disclosed herein and/or arrays of such devices. Other applications include watch crystals, automotive and architectural glass including sunroofs and privacy glass. The photovoltaic devices may be used for active power generation, e.g., for completely self-powered applications and battery charging (or battery life extension).

(62) In conclusion, near-infrared absorbing, transparent planar organic solar cells with a maximum power of 1.40.1% and average visible transmission of exceeding 55.+0.2% have been demonstrated. This average visible transmission is sufficiently transparent for incorporation on architectural glass. The excitonic character of organic semiconductors is advantageously exploited to produce unique photovoltaic architectures not easily accessible via inorganic semiconductors. By positioning the active layer absorption selectively in the NIR, it is possible to optimize the architecture using a NIR reflector composed of a DBR mirror centered at 800 nm that results in a transparent solar cell efficiency approaching that of the non-transparent control cell. Ultimately these devices provide a guide for achieving high efficiency and high transparency solar cells that can be utilized in windows to generate power, reduce cooling costs, and scavenge energy in a variety of applications.