ENVIRONMENTAL ASSESSMENT OF TRANSPARENT PHOTOVOLTAIC S By Eunsang Lee A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Environmental Engineering Œ Doctor of Philosophy 2018 ABSTRACT ENVIRONMENTAL ASSESSMENT OF TRANSPARENT PHOTOVOLTAICS By Eunsang Lee Transparent organic photovoltaics (TPV) can harvest energy from the near -infrared (NIR) and ultraviolet region of the solar spectrum and could be used in new applications such as windows. In addition to producing electricity , the transparent organic solar cell absorbs in the NIR region and could reduce the cooling energy demand of building s during summer. Organic photovoltaic (OPV) is an emerging technology, developed mostly to replace fossil fuel energy , aimed to reduce greenhouse gases emissions. As OPV technology is reaching commercialization, it is essential to quantify its environmental impact s and ensure that new issues are not created . Life cycle assessment (LCA) is often used to compare energy technologies and identify environmental concerns , but this process is challenging for emerging technologies due to lack of inventory data. To guide future transparent OPV development , this work (1) demonstrates a new iterative methodology to evaluate and guide OPV material manufacturing that combines LCA and gr een chemistry approaches, (2) evaluate s the energy saving from organic TPV in window and skylight applications in various cities, and (3) assess es the impact of organic TPV on urban heat island effect. The methodology was used to identify fihotspot s,fl which correspond to the process es that have the highest impact for chloroaluminum phthalocyanine (ClAlPc) manufacturing . An optimized process that reduces the environmental impact by 3%, the cost by 9% and chemicals hazard by 23% compared to the current process was demonstrated. The impact of TPV during the use phase was studied using ClAlPc based devices in window application. The building energy performance was shown to improv e by up to 20 % due to heating and cooling energy saving. The energy saving var ies with climate since NIR absorption by TPV in a window is more efficient in a warmer climate. The use of TPV in the window application in an urban area could further reduce the energy demand of buildings. The net energy saving by the TPV application in the urban area was calculated to be higher than in rural area by up to 2 GJ per month. iv ACKNOWLEDGEMENTS First of all , I would like to thank my advisor Dr. Annick Anctil for her motivation and guidance. She has motivated and guided me to become an independent researcher. Without her supports, I would not be able to complete my research. I also want to appreciate my comm ittee members, Dr. Richard Lunt, Dr. Volodymyr Tarabara , and Dr. Mehrnaz Ghamami for their helpful suggestions on my research. Special thanks to Dr. Richard Lunt for allowing me to attend his group meeting and learn knowledge on organic photovoltaics. I wa nt to acknowledge all my colleagues in Michigan State University for sharing their expertise with me and support my research. Lastly, I would like to thank my family for their encouragements. v TABLE OF CONTENTS LIST OF TABLES ........................................................................................................................ vii LIST OF FIGURES ....................................................................................................................... ix KEY TO ABBREVIATIONS ....................................................................................................... xii Chapter 1 Introduction .................................................................................................................... 1 1.1 Introduction ............................................................................................................... 1 1.1.1 Motivation .......................................................................................................... 5 1.2 Background ............................................................................................................... 7 1.2.1 Green Chemistry ................................................................................................. 7 1.2.2 Life Cycle Assessment ....................................................................................... 8 1.2.3 Transparent Organic Photovoltaic Applications ................................................ 9 Chapter 2 Alternative Chl oroaluminum Phthalocyanine Synthesis .............................................. 11 2.1 Alternative Synthesis Pathway ................................................................................ 11 Chapter 3 Evaluation of ClAlPc Synthesis Methods for Transparent Organic Photovoltaic ....... 17 3.1 Introduction ............................................................................................................. 17 3.2 Experimental Procedure .......................................................................................... 18 3.2.1 Material Synthesis ............................................................................................ 18 3.2.2 UV -visible Spectroscopy .................................................................................. 19 3.2.3 HPLC -MS Spectroscopy .................................................................................. 19 3.2.4 Cost Analysis .................................................................................................... 20 3.2.5 Solar Cell Fabrication and Tests ....................................................................... 20 3.3 Result and Discussion ............................................................................................. 21 3.3.1 UV -visible Spectra ........................................................................................... 21 3.3.2 HPLC -MS Spectra ............................................................................................ 21 3.3.3 Cost Analysis .................................................................................................... 22 3.3.4 Photovoltaic Performance ................................................................................ 23 3.4 Conclusion and Future Work .................................................................................. 25 Chapter 4 Fine Chemical Process Toward to Sustainable Organic Photovoltaic ......................... 26 4.1 Introduction ............................................................................................................. 26 4.2 Method .................................................................................................................... 29 4.2.1 Life Cycle Assessment for Fine Chemicals Process ........................................ 29 4.2.2 Selection of Impact Category for Sustainability Assessment ........................... 32 4.2.3 Material Synthesis and Data Collection ........................................................... 34 4.3 Result and Discussion ............................................................................................. 36 4.3.1 Methodology Assessment ................................................................................. 36 4.3.2 Overall Evaluation ............................................................................................ 41 Chapter 5 Building Energy Model for Life Cycle Assessment of Transparent Organic Photovoltaic in Window Application ................................................................................................................. 44 vi 5.1 Introduction ............................................................................................................. 44 5.2 Method .................................................................................................................... 46 5.2.1 Selection of study area ...................................................................................... 48 5.2.2 Simulation for Building Energy Demand and Electricity Production .............. 50 5.3 Result and Discussion ............................................................................................. 54 5.3.1 Simulation for Building Energy Demand and Photovoltaic ............................. 54 5.3.2 Assessment of Building Energy Demand by TPV Application ....................... 56 Chapter 6 TPV Application in Urban Area: Urban Heat Island Effect and Its Consequential Environmental Impact on Building Sector ................................................................................... 59 6.1 Int roduction ............................................................................................................. 59 6.2 Method .................................................................................................................... 61 6.3 Result and Discussion ............................................................................................. 63 6.3.1 Comparison of Urban and Rural Temperature ................................................. 63 6.3.2 UHI Intensity in Urban Area and Waste Heat from HVAC System ................ 65 6.3.3 Energy Saving by TPV Application in Rural and Urban Areas ....................... 68 Chapter 7 Conclusions .................................................................................................................. 70 APPENDIX ................................................................................................................................... 73 REFERENCES ............................................................................................................................. 90 vii LIST OF TABLES Table 1 -1. Twelve principles of green chemistry and description (Adapted from Mulvihill, 2011) ......................................................................................................................................................... 8 Table 2 -1. Summary of reaction energy source, conventional and microwave reactors: principle, advantages, and disadvantages ...................................................................................................... 15 Table 3 -1. Sample description of chloroaluminum phthalocyanine synthesis from phthalonitrile and phthalic anhydride .................................................................................................................. 19 Table 3 -2. Photovoltaic performance parameters of optimal chloroaluminum phthalocyanine devices: ClAlPc -Reference (TCI America), phthalonitrile (PN -1), and phthalic anhydride (PA) cells ............................................................................................................................................... 24 Table 4 -1. Summary of green chemistry metrics and LCA impact categories used in the assessment methodology and to guide alternative s ynthesis ........................................................................... 31 Table 4 -2. Reaction condition and chemical use for ClAlPc processes ....................................... 36 Table 5 -1. Simulation inputs for building energy demand and TPV application ........................ 51 Table 5 -2. Thermal properties of envelope used for DOE reference building in each study area 52 Table 5 -3. Heating source of the office building in study area assessed from CBECS database 54 Table 6 -1. Urban Weather Generator Parameters as the inputs of simulation for generating urban temperature profiles ...................................................................................................................... 62 Table A1. Summary of existing method for fine chemical evaluation ........................................ 75 Table A2. Data sources for materials and energy used in the inventory analysis ........................ 85 Table A3. Chemical identification and NFPA score of mater ials used in ClAlPc processes ...... 86 Table A4. Chemical hazard weighed to material consumption for ClAlPc manufacturing based on input from technosphere ............................................................................................................... 87 Table A5. Chemical hazard weighed to material consumption for ClAlPc manufacturing based on input from tech nosphere except energy source (grey in Table A2.) ............................................. 88 Table A6. Green chemistry metrics of proposed processes and solvents use .............................. 88 Table A7. Reference cost of crude oil, coals, and natural gas ..................................................... 89 Table A8. Production rate of baseline and alternative processes ................................................. 89 viii Table A9. Summary of sustainability scores for all methods ...................................................... 89 ix LIST OF FIGURES Figure 1 -1. Net electricity generation from select fuels (Adapted from Annual Energy Outlook 2018, EIA, 2018) ............................................................................................................................. 2 Figure 1 -2. Ren ewable electricity generation, including end -use generation (Adapted from Annual Energy Outlook 2018, EIA, 2018) .................................................................................................. 2 Figure 1 -3. Energy use for power generation and total electricity demand in 2010 (Adapted from Transition to Sustainable Buildings, IEA, 2013) ............................................................................ 3 Figure 1 -4. (a) Cradle to grave Life Cycle Assessment and (b) LCA framework according to ISO standards 14040 and 14044(Adapted from Anctil and Fthenakis, 2011). © 2011 Annick Anctil and Vasilis Fthenakis. Originally published in Life Cycle Assessm ent of Organic Photovoltaics under CC by 3.0 license. Available from: DOI: 10.5772/38977 .............................................................. 9 Figure 2 -1. (a) Synthesis pathway of chloroaluminum phthalocyanine from phthalonitrile and phthalic anhydride (b) reaction condition and reactants of chloroaluminum phthalocyanine ...... 12 Figure 2 -2. Synthesis pathway of chloroaluminum phthalocyanine via phthalonitrile showing atom utilization of reactants .......................................................................................................... 13 Figure 2 -3. Synthesis pathway of chloroaluminum phthalocyanine via phthalic anhydride showing atom utilization of reactants .......................................................................................................... 13 Figure 2 -4. Reaction energy input of chloroaluminum phthalocyanine synthesis from phthalonitrile under microwave reactor and pictures of the synthesis at each reaction stage (1. initial mixture of reactants, 2. intermediate reaction, 3. chloroaluminum phthalocyanine in 1 -chloronaphthalene) ........................................................................................................................ 14 Figure 2 -5. Reaction energy input of chloroaluminum phthalocyanine synthesis from phthalic anhydride under microwave reactor and pictures of the synthesis at each reaction stage (1. initial mixture of reactants, 2 -4. intermediate reactions, 6. chloroaluminum phthalocyanine in 1-chloronaphthalene) ........................................................................................................................ 14 Figure 2 -6. Flow diagram of chloroaluminum phthalocyanine processes: (a) reference process using a heatin g mantle and (b) alternative process using a microwave reactor ............................ 16 Figure 3 -1. Metal phthalocyanines synthesis from (1) phthalonitrile (PN) and (2) phthalic anhydride (PA) .............................................................................................................................. 18 Figure 3 -2. Energy level diagram and structure for the solar cell based on chloroaluminum phthalocyanine and C 60 ................................................................................................................. 20 Figure 3 -3. UV-vis spectrum of chloroaluminum phthalocyanine: ClAlPc -reference (green solid), ClAlPc -phthalonitrile (blue dash -dot), and ClAlPc -phthalic anhydride (red dash) ...................... 21 x Figure 3 -4. Mass chromatogram (left) and isotope distribution (right) at 580.16 m/z of chloroaluminum phthalocyanine: ClAlPc -Reference (green), ClAlPc -Phthalonitrile (blue ), ClAlPc -Phthalic anhydride -1 (deep red) and Phthalic anhydride -2 (red) ..................................... 22 Figure 3 -5. Cost analysis of chloroaluminum phthalocyanine: ClAlPc from phthalic anhydride (PA process) and phthalonitrile (PN process) compared to purchased ClAlPc from TCI America ....................................................................................................................................................... 23 Figure 3 -6. Representative current density -voltage characteristics of solar cells based on ClAlPc -reference (green solid), ClAlPc -phthalonitrile -1 best (blue dash -dot) and phthalonitrile -1 replication (deep blue dash) and ClAlPc -phthalic anhyd ride (red dot) ........................................ 24 Figure 3 -7. External quantum efficiency of solar cells based on ClAlPc -reference (green solid), ClAlPc -phthalon itrile -1 best (blue dash -dot) and phthalonitrile -1 replication (deep blue dash) and ClAlPc -phthalic anhydride (red dot) ............................................................................................. 25 Figure 4-1. (a) Proposed iterative methodology and (b) sustainability method and categories ... 29 Figure 4 -2. (a) Summary of hotspots deter mined by iterative evaluation, and (b) -(e) evaluation of alternative ClAlPc processes based on environmental, chemical hazard and cost impacts .......... 37 Figure 4 -3. Cumulative energy demand for baseline, 1 -chloronaphthalene (P1), 2,4 -dichloroanisole (P4), and diethylene glycol (P5) ClAlPc processes ............................................. 42 Figure 4 -4. Sankey diagram of the cumulative energy flow for ClAlPc synthesis from (a) the baseline and (b) diethylene glycol and DBU (P5) excluding transport and electricity ................ 43 Figure 5 -1. System diagram shows: (a) scope of simulations in building system simulation and photovoltaic simulatio n for TPV application in window, (b) system boundaries of life cycle assessment including material production and module manufacturing, and module use of TPV as windows and skylights, (c) scenarios of TPV application. baseline window composites of 3mm glas s and 13 mm air gap; glass TPV that contains the active layer of TPV represents window replacement; Plastic TPV encapsulated with PET film is a scenario of retrofitting window. ..... 47 Figure 5 -2. Annual climate conditions (global horizontal irradiance, relative humidity, and dry bulb temperature) of study areas ................................................................................................... 48 Figure 5 -3. Summary of study areas based on the electricity grid and climate zone. eGrid is a source of data on the environmental characteristics of electric power generated in the U.S. Thermal criteria classify IECC climate zone based on degree days. .......................................................... 49 Figure 5 -4. Optical properties of transparent photovoltaic as input of EnergyPlus simulation: (a) front illumination (b) back illumination ....................................................................................... 53 Figure 5 -5. Monthly energy demand for baseline building and electricity production by TPV application. Baseline building is installed double pane window, and power conversion efficiency of TPV for electricity production is 10 %. .................................................................................... 55 xi Figure 5 -6. Annual energy saving of TPV application in window for medium office: (a) glass TPV (b) plastic TPV, with application of TPV in SEWN (south, east, west, and north) direction ...... 56 Figure 5 -7. Annual energy saving of TPV application in skylight for sma ll office (a) glass TPV (b) plastic TPV .............................................................................................................................. 57 Figure 6 -1. Description of Urban weather Generator (UWG) model, and adoption of DOE reference building applied TPV in window (model mechanics are adapted from http://urbanmicroclimate .scripts.mit.edu). Reference weather data from the weather station is simulated to generate urban weather data by reflecting urban characteristics including land use, wind speed, a reflection of solar radiation, traffic, and waste heat from the building. ................. 61 Figure 6 -2. Temperature variation in Los Angeles based on rural scenario (green dot line) and urban: baseline ( solid black line) and urban: TPV (red dash line) scenarios during the summer season for a month (Aug.) ............................................................................................................. 64 Figure 6 -3. Temperature variation in Los Angeles based on rural scenario (green dot line) to urban: baseline (solid black line) and urban: TPV (red dash line) scenarios during the winter season for a month (Dec.) ................................................................................................................................. 65 Figure 6 -4. Urban heat island intensity in Los Angeles during (a) summer season (Aug.) and (b) winter season (Dec.) ...................................................................................................................... 66 Figure 6 -5. Emission of waste heat from HVAC to the urban area in (a) summer season (Aug.) (b) winter season (Dec.) ...................................................................................................................... 68 Figure 6 -6. Comparison of monthly energy saving for TPV application in window for urban and rural scenarios in Los Angeles during summer (Aug.) and winter (Dec.) seasons ....................... 69 Figure A1 . Flow diagram of the baseline process ........................................................................ 78 Figure A2. Flow diagram of P1 ................................................................................................... 78 Figure A3. Flow diagram of P2 ................................................................................................... 79 Figure A4. Flow diagram of P3 ................................................................................................... 79 Figure A5. Flow diagram of P4 ................................................................................................... 80 Figure A6. Flow diagram of P5 ................................................................................................... 81 Figure A7. UV-vis spectra of ClAlPc samples ............................................................................ 83 Figure A8 . HPLC -MS spectra of ClAlPc samples (at 6.81 min) ................................................. 84 xii KEY TO ABBREVIATIONS 1-ClNP : 1-chloronaphthalene 2,4-DCA: 2,4-dichloroanisole AE: Atom economy AlCl 3: Aluminum chloride anhydrous BCP: Bathocuproine BIPV : Building integrated photovoltaic BOS: Balance -of-systems C60: C60 fullerene CBECS : Commercial Buildings Energy Consumption Survey CED : Cumulative energy demand ClAlPc : Chloroaluminum phthalocyanine CuPc : Copper phthalocyanine DBU: 1,8-Diazabicyclo[5.4.0]undec -7-ene DEG : Diethylene glycol eGrid : emissions & Generation Resource Integrated Database EIA : U.S. Energy Information Administration EPBT : Energy payback time EQE : External quantum efficiency EROI : Energy return on energy invested GHGs: Green housegases GHS: Globally Harmonized System GWP : Global warming potential ITO : Indium tin oxide LCA : Life cycle assessment LCI : Life cycle inventory MoO 3: Molybdenum(VI) oxide M-Pcs : Metallophthalocyanines NEB : Net environmental benefit NH4Cl: Ammonium chloride NIR : Near -infrared NOx: Nitrogen oxides OPV : Organic photovoltaics PA: Phthalic anhydride PCE : Power conversion efficiency PM: Particulate matter PMI : Process mass intensity PN: 1,2-dicyanobenzene PV: Photovoltaic QSAR : Quantitative structure -activity relationship models SHGC : Solar heat gain coefficient SO2: Sulfur dioxide SRR : Skylight to roof ratio TMY : Typical Meteorological Year xiii TPV: Transparent OPV UHI: Urban heat island effect UV: Ultraviolet UWG : Urban weather generator WDI : Water demand indicator WWR : Window to wall ratio 1 Chapter 1 Introduction 1.1 Introduction Future energy solutions should not create new environmental, social, and cost problems . Electricity generation from fossil fuels, such as coal, natural gas, and fuel oils , is not desirable for a long -term goal. Environmental impacts related to these non -renewable sources of electricity include the release of a large amount of greenhouse gases (GHGs ).1 To solve the issue, interest in renewable energy has inc reased, and the most promising renewable source is solar energy because of over hundred thousand terawatts of potential energy from sunlight. 2 Advancing photovoltaic (PV ) technology over time has result ed in interest for alternative photovoltaic s, such as organic photovoltaics (OPV ). OPV are soon to be commercialize d because of their benefits such as flexibility, low weight, and low cost. 3,4 Currently, electricity generat ed in the U.S. relies on the use of coal and natural gas . Also , coal -fired electric ity generation increase s air pollution due to mercury, sulfur dioxide (SO 2), nitrogen oxides (NOx ), and particulate matter (PM ) emissions .5 The next fifty years are critical in order to replac e conventional energy sources to renewable energy in order to reduce the environmental impact of electricity production . According to the U.S. Energy Information Administration (EIA ), the net coal -fired electric ity generating capacity has decreased by around 60 gigawatts (GW) from 2011 to 2016 to comply with the U.S. Environmental Protection Agency™s Mercury and Air Toxics Standards. 6 Electricity generation from renewable energy should reach 1,650 billion kilowatt -hours (BkWh) by 2050 and should become the second largest source of electricity as shown in Figure1 -1.6 2 Figure 1-1. Net electricity generation from s elect fuels (Adapted from Annual Energy Outlook 2018, EIA, 2018) 6 Figure 1-2. Renewable electricity generation, including end -use generation (Adapted from Annual Energy Outlook 2018, EIA, 2018) 66 The amount of renewable energy produced from wind, geothermal , and hydroelectricity are expected to remain similar while PV would inc rease until 2050 as shown in Figure 1 -2.6 Interest in b uilding integrated photovoltaic (BIPV ) application has increase d due to its potential economic 3 and environmental benefits. 7 Reducing land use for PV deployment is a major economic benefit of BIPV in an urban area, and BIPV application could also achieve microclimate improvement. 7 Building sector contribute d to over 50% of total energy use in high -income cities (Tokyo, New York, Seoul, Greater London, Singapore, Berlin, and Bologna) in 2008 . The energy use in the building sector was associated with building maintenance and operation. 8 Buildings are complex systems , and their energy consumption is affected by local climate, occupancy profiles, and consumer preferences. 9 Figure 1 -3 shows the primary energy use for electricity generation and final world electricity demand for each sector . A key solution for urban sustainable energy is to increase the energy efficiency of the building energy system and/ or to increase low carbon electricity production in a city where land is limited, which can mostly be attained using BIPV. Figure 1-3. Energy use for power generation and total electricity demand in 2010 (Adapted from Transition to Sustainable Buildings, IEA, 2013) 9 Although it is difficult to estimate, windows are most likely responsible for 5% to 10% of the total energy consumed in buildings in OECD countries. 10 It is essential to develop high -performance windo ws with low thermal transmittance and climate -appropriate solar heat gain coefficient (SHGC ) to reduce energy consumption . Efforts have focused on multiple glazed 4 windows (triple or quadruple) with multiple low -E coatings, exotic inert gases, such as krypton and xenon to develop high -performance windows. However, these windows are generally too expensive , and additional work is required to find an inexpensive solution to reduce the energy consumption of windows .10 Transparent OPV (TPV ) offer s an opportunity to produce electricity and improve building energy performance at the same time. The absorption from TPV can be tuned from ultraviolet (UV , <435 nm) to near infrared (NIR , >670 nm) range of t he solar spectrum . TPV in window applications could control the inward flow of solar heat and reduce cooling energy demand while the absorbed solar energy can be used to produce electricity while performing the original function of windows because of its t ransparency .11 Although there a re benefits of OPV applications, the potential impacts from OPV material manufacturing are not well understood . Previous work has used life cycle assessment (LCA ) to evaluate the impact associated with fullerene s synthesi s12, a common acceptor material in OPV as well as various small molecules and polymer organic solar cells 3 based on cumulative energy demand (CED ). The energy requirement for OPV manufacturing was 50 % lower than for manufacturing conventional inorganic photovoltaics .3 While the potential energy benefit of OPV compared to inorganic solar cells has been established , the imp act of transparent photovoltaics has not been studied . One limitation of LCA i n its application to emerging technology which uses new chemicals is the lack of inventory information . An approach to reduce the environmental and health impacts of chemical s and help develop alternative chemical process is green chemistry. The framework of green chemistry include s all stages of the chemical life -cycle and tries to reduce the intrinsic hazard of chemical products by design. 13 Applying green chemistry to the new chemical is relativ ely 5 simple compared to LCA because gree n ch emistry metric is designed to use s toichiometry or mass balance to evaluat e envir onmental performance. 14 1.1.1 Motivation Energy solutions developed today should remain desirable in the future. As organic solar cells are getting closer to commercialization, it is imperative to investigate the potential unintended consequences in both the manufacturing and use phase s of the organic solar cell s. LCA considers all stage s from raw material extraction (cradle) to the ultimate return of the material to the earth (grave). The cumulative environmental impacts from all stages in the product are imperative since they allow the evaluation of trade -offs in product and process selection. Performing LCA during the development process can lead to design solutions with lower environmental , cost, and health impact s. Sustainability considers the economic, social, and environm ental aspects altogether . Developing technology for only photovoltaic efficiency without considering embodied energy and environmental impacts should be avoided. Building renewable energy infrastructure requires environmental costs to help choose appropria te materials for photovoltaic technologies .15,16 It is also better to use environmentally benign chemicals and reduce the amount of waste during the manufacturing of OPV material , since prevention is better than remediation. A holistic approach to evaluat ing the sustainability of OPV material process has not been established , but it could play a significant role as the first step for maximizing net benefit of OPV applicati on. TPVs have the potential to be affordable by using low -cost materials and to be integrat ed into the infrastructure of the built environment , thereby reduc ing installation costs and balance -of-systems (BOS ).11 Currently, the assessment o f the environmental performance of PV is limited to 6 the traditional function and application of PV by only considering the module, BOS, and electricity generation. The tradeoff between PV efficiency and energy inputs, such as cumulative energy demand (CED) in per spective of energy performance , has been studied and formulated as metrics such as energy payback time (EPBT ) and energy return on energy invested (EROI ).17 However, TPV generates electricity and contributes to management of building energy demand by NIR absorption . The NIR absorption by TPV has not been well evaluated . However , switching from a conventional double -pane window to a semi -transparent PV made o f amorphous silicon as a window has been found to reduce the building energy demand up to 15 % .18 Integrating the benefit from TPV in the window application to current m ethods of PV evaluation is required to make a clear comparison with existing renewable energy application s. Combining the energy saving and the electricity production to evaluate the benefit of TPV is needed to compare with other renewable energy options . Urban area s account for about 64 % of the global primary energy use and produce 70 % of the world carbon dioxide emission s.19 Serious action to reduce the energy burden needs to be done , and some cities such as Aspen, Colorado, and Burlington, Vermont already use only renewable energy . Other s, such as San Diego, California, aim to be 100% powered by renewable sources by 2035.19 The use of TPV in window applicatio n could help fulfill this goal of renewable energy in the urban area. An additional impact of energy use in the urban area is due to waste heat from anthropogenic activities and solar radiation which cause urban heat island effect (UHI ). UHI can potentially affect communities by increasing energy demand during summertime peak, air conditioning costs, air pollution , greenhouse gas emissions, etc. 20 Large -scale deployment of BIPV in urban area s, such as rooftop or façade application, has been studied for microclimate impact , and studies conclude that PV application could reduce local temperature because solar energy is 7 converted to electric ity 21,22 As a similar application, s tudying the impact of TPV in urban area is required to e nsure that TPV does not further increase UHI effects. 1.2 Background 1.2.1 Green Chemistry Green chemistry is known for its twelve principles (summarized in Table 1 -1) and metrics 13 which are general guidelines to reduce the environmental impact of chemical synthesis . The fra mework of green chemistry requires to (1) consider all stages of the chemical life -cycle, (2) reduce the intrinsic hazard of chemical products by design, and (3) use green chemistry as a cohesive system of principles or design criteria. 13 As an example of green chemistry metrics and applications, the E-factor (Equation 1 -1) is used to describe fine chemical (chemical category of OPV materials) processes , due to a large amount of waste generated from solvent use. 23 Similarly, a pharmaceutical company report ed that the wasted materials were mainly from solvent use as reaction medium or work -ups. The contribut ion from solvent use was up to 35 to 50 % material use in a pharmaceutical process .24 This issue is the motivation for searching for green solve nts that reduce environmental impacts on fine chemical and pharmaceutical manufacturing 25, and most pharmaceutical companies have adopt ed green chemistry app roach .26 Efactor = () (1-1) Although there are many studies reporting efforts on adopting green chemistry approaches in the pharmaceutical area, green chemistry applications in OPV area have been little reported . Previous study discusses strategies for producing conjugated polymers using green chemistry 8 toward sustainable OPV. 15 This paper criticiz es the current research focused on renewable energy that consideration of embodied energy and environmental impacts is often absent . It further suggests guidelines to produce little waste, avoid protecting groups, use catalysts, generate environmentally benign byproducts, and consume less energy with exemplary polymer synthesis pathways. 15 Table 1-1. Twelve principles of green chemistry and description (Adapted from Mulvihill, 2011) 27 No. Principle Description 1 Prevention It is better to prevent waste than to treat or clean up waste after it has been created 2 Atom eco nomy (AE) Synthetic methods should be designed to maximize the incorporation of all materials used in the process into the final product 3 Less hazardous chemical syntheses Wherever practicable, synthetic methods should be designed to use and generate sub stances that possess little or no toxicity to human health and the environment 4 Designing safer chemicals Chemical products should be designed to effect their desired function while minimizing their toxicity 5 Safer solvents and auxiliaries The use of auxiliary substances (e.g., solvents, separation agents, and others) should be made unnecessary wherever possible and innocuous when used 6 Design for ene rgy efficiency Energy requirements of chemical processes should be recognized for their environmental and economic impacts and should be minimized . If possible, synthetic methods should be conducted at ambient temperature and pressure 7 Use of renewable feedstocks A raw material or feedstock should be renewable rather than depleting whenever technically and economically practicable 8 Reduce derivatives Unnecessary derivatization (use of blocking gr oups, protection/deprotection, temporary modification of physical/chemical processes) should be minimized or avoided if possible because such steps require additional reagents and can generate waste 9 Catalysis Catalytic reagents (as selective as possible) are superior to stoichiometric reagents 10 Design for degradation Chemical products should be designed so that at the end of their function they break down into innocuous degradation products and do not persist in the environment 11 Real -time a nalysis for pollution prevention Analytical methodologies need to be further developed to allow for real -time, in -process monitoring and control prior to the formation of hazardous substances 12 Inherently safer chemistry for accident prevention Substance s and the form of a substance used in a chemical process should be chosen to minimize the potential for chemical accidents, including releases, explosions, and fire 1.2.2 Life Cycle Assessment Life cycle refers to the lifespan of a product and/or process and include s raw material extraction, material processing and manufacturing, assembly, product use, and end of life as shown in Figure 1 -7(a). 28 LCA is a systematic approach and consists of goal definition and scope, inventory analysis, and impact assessment, and interpretation as illustrated in Figure 1 -7(b). 28,29 9 Figure 1-4. (a) Cradle to grave Life Cycle Assessment and (b) LCA framework according to ISO standards 14040 and 14044(Adapted from Anctil and Fthenakis, 2011) .28 © 2011 Annick Anctil and Vasilis Fthenakis. Originally published in Life Cycle Assessment of Organic Photovoltaics under CC by 3.0 license. Available from: DOI: 10.5772/38977 LCA can be applied to determine the potential environmental impacts from any product, process, or service. The accuracy and detail of the data collected during the life cycle inventory (LCI ) are important and will determine the accuracy and quality of the analysis. The life cycle impact assessment (LCIA) identifies a link between the product or process and its potential impacts. Finally, interpretation evaluate s the results of LCI and LCIA and communicate the results effectively .29 1.2.3 Transparent Organic Photovoltaic Application s Technologies of TPV have recently emerged with excitonic materials that selectively harvest electricity from UV and NIR radiation. Materials used to fabricate TPV include organic small molecules and organic salts. 30,31 For TPV device s with small molecule s, NIR -selective TPV was demonstrated in 2011 with chloroaluminum phthalocyanine (ClAlPc ) as the donor and C60 fullerene ( C60) as the acceptor between two transparent conductive electrodes (indium tin oxide, ITO). The power conversion efficiency (PCE ) of the demonstrated device was 1.3± 0.1% with average visible transmittance ( AVT) of 65% .30 Theoretical PCE of TPV is 20.6 % with 100% AVT 10 for a single junction device , and could reach about 30% PCE and 100% AVT with three to four junctions devices .11 The performance of small molecule OPV relies on material purity. A study reports that PCE of solar cell made of copp er phthalocyanine (CuPc ) varies from 0.26 ± 0.01 % to 1.4 ± 0.1 % based on the purity of CuPc. 32 The purification of OPV material is critical to the embodied energy of OPV device, which is an evaluating criterion of environmental performance of PV. For example, the purification of C 60 to reach electronic grade purity doubles the embodied energy .12 11 Chapter 2 Alternative Chloroaluminum Phthalocyanine Synthesis 2.1 Alternative Synthesis Pathway Chloroaluminum phthalocyanine (ClAlPc) is of particular interest for TPV since it absorbs only in the near -infrared region. It allows the solar cell to be transparent, which is important for applications such as power -generating windows. 11,30 While metal phthalocyanines are common small molecules widely used in the dye industry, 33 the required high purity for solar applications 32 makes it a fine chemical. The most common prec ursors of phthalocyanine are phthalic anhydride (PA ) and 1,2 -dicyanobenzene (phthalonitrile, PN ). PN can be produced from o-xylene under ammoxidation process at 480 . The yield is 80-85%.34 The other precursor s are metal salts, and aluminum c hloride anhydrous (AlCl 3) is used to synthesis ClAlPc. Manufacturing process of AlCl 3 is by chlorinating molten aluminum at 670 -850°C, and the yield is over 90 %. 35 The reaction of ClAlPc from PN and PA is carried out using either heating mantle or microwave reactor (CEM Corporation Discover SP). In C lAlPc synthesis, the molar ratio of PN or PA to aluminum ion is fixed to 4:1 and mixed in reaction media .33,36,37 High boiling point solvent is required for all synthesis because of t Synthesizing methods of ClAlPc are reported in limited number of publication s since the material is not typically utilized in the dye industr y which consume s most metallophthalocyanines (M-Pcs ).38Œ40 The summary of the synthesis pathway of ClAlPc and reaction condition is shown in Figure 2 -1,33,36,37 and is used to calcu late atom economy (AE ) that evaluat es the environmental performance of the synthesis pathway bas ed on stoichiometr y.41 12 Figure 2-1. (a) Synthesis pathway of chloroaluminum phthalocyanine from phthalonitrile and phthalic anhydride (b) reaction condition and reactants of chloroaluminum phthalocyanine As a screening process, the environmental performance of the synthesis pathway of ClAlPc is evaluated using green chemistry metrics. AE is the conversion efficiency of a chemical process considering all atoms in the reactants and the desired products produced as shown Equation 2 -1. (2-1) Synthesis pathways of ClAlPc from either PN or PA showing atom utilization of reactants are shown in Figure. 2 -2 and Figure. 2 -3. The pathway from PN results in higher AE becau se most reactants are converted to ClAlPc while PA pathway use s urea only as a nitrogen source (Figure 2-3, orange color), and residual atoms are wasted . Therefore, ClAlPc synthesis via PN would be desirable in term of environmental performance so that fur ther optimization of ClAlPc process starts from PN pathway. (%)= . ( . )×100 13 Figure 2-2. Synthesis pathway of chloroaluminum phthalocyanine via phthalonitrile showing atom utilization of reactants Figure 2-3. Synthesis pathway of chloroaluminum phthalocyanine via phthalic anhydride showing atom utilization of reactants As evidence of byproduct formation during ClAlPc synthesis from PN and PA, the energy input s from a microwave reactor and images at certain temperature s are recorded . Figure 2 -4 shows continuous energy input until around 80 °C, and then intermediate s start to form ClAlPc with an exothermic reaction that is recognized with the trend of cumulative energy input. Meantime, ClAlPc synthesis from PA shows the opposite trend of cumulative energy input compared to the ClAlPc synthesis from PN. The images show color changes from orange to green and eventually purple. Bubbles are observed during temperature increase and are suspected to be hydrochloric (%)=574.964×128.13+133.34×100=89.02% (%)=574.964×148.12+4×60.06+133.34×100=59.52% 14 acid gas, due to the reaction of aluminum chloride with moisture. 42 The formation of hydrochloric acid gas results in loss of chlorine ion during reaction , so the yield of ClAlPc from PA is only around 25 %. Figure 2-4. Reaction energy input of chloroaluminum phthalocyanine synthesis from phthalonitrile under microwave reactor and p icture s of the synthesis at each reaction stage (1. initial mixture of reactants, 2. intermediate reaction, 3. chloroaluminum phthalocyanine in 1 -chloronaphthalene ) Figure 2-5. Reaction energy input of chloroaluminum phthalocyanine synthesis from phthalic anhydride under microwave reactor and p icture s of the synthesis at each reaction stage (1. initial mixture of reactants, 2 -4. intermediate reactions, 6. chloroaluminum phthalocyanine in 1 -chloronaphthalene ) 15 The source of reaction energy is of interest from the green chemistry perspective , which corresponds to #6 design for energy efficiency from Table 1-1. M-Pcs have been synthesized using microwave reactor, and the reaction time is five to ten min 43Œ45. The experiments require d 40 minute which is longer than other M -Pcs but significantly shorter than six hours when using the heating mantle method 46. For microwave synthesis , polar solvents are required because heating is based on molecular rotation 47, and the comparison of con ventional and microwave reactor is shown in Table 2 -1. Table 2-1. Summary of reaction energy source, conventional and microwave reactors : principle, advantage s, and disadvantage s Current ClAlPc process es use toxic and hazardous solvents, and thus there are opportunities to reduce health and environmental impacts. Extra filtration and solvent reflux are minimized to reduce the energy input from reference method. 46 The similar boiling point of toluene and D.I. water interrupt s with solvent recovery and result s in more energy consumption for solvent 16 recovery . Acetone is used as a substitute of toluene because it has low er boiling point , reducing the energy consumption for solvent recovery. The reference and alternative processes are shown in Figure 2 -6. Trichlorobenzene is the most common reaction media in copper phthalocyanine synthesis, but production of polychlorinated biphenyl as a to xic byproduct has been reported .37,48,49 ClAlPc process uses 1 -chloronaphthalene which has the same concern and should be replaced . Figure 2-6. Flow diagram of chloroaluminum phthalocyanine processes : (a) reference process using a heating mantle 46 and (b) alternative process using a microwave reactor 17 Chapter 3 Evaluation of ClAlPc Synthesis Methods for Transparent Organic Photovoltaic 3.1 Introduction Metal phthalocyanines (M -Pcs) have been widely used as electron d onor material in organic photovoltaic (OPV). 50 Especially, chloroaluminum phthalocyanine (ClAlPc) is of particular interest for OPV since it absorbs only in the near -infrared (NIR) region. 30,51 Œ54 It allows the solar cell to be transparent which is important for applications such as power -generating windows. For this type of application, the merit of the solar cell is not solely a function of device efficiency but also transparency. The focus of this research is ClAlPc for which there is limited reported synthesis methods since the material is not typically utilized in the dye industry that consumes most M -Pcs. 39,40 Two precursors have been mainly used to synthesis M-Pcs: (1) phthalonitrile (PN) and (2) phthalic anhydride (PA), as shown in Fi gure3 -1. Although it is known that PN produces higher purity M -Pcs, PA is preferred for copper phthalocyanine (CuPc) synthesis by the dye industry because of low cost , phthalic anhydride being about 100 times cheaper than phthalonitrile. 37 However, the PA process increases impurity in M -Pcs, which reduces the efficiency in OPVs. A previous study showed that power conv ersion efficiency of donor -acceptor bilayer solar cell based on CuPc increase from 0.26 ± 0.01% to 1.4 ± 0.1% with increasing purity. 32 The same study also identified metal -free phthalocyanine as a potential impurity in CuPc using mass spec trometry which reduces hole mobility and device efficiency. 32 18 Figure 3-1. Metal phthalocyanines synthesis from (1) phthalonitrile (PN) and (2) phthalic anhydride (PA) For the PN process, phthalonitrile directly forms chloro -substituted CuPc, which may impact the efficiency in OPVs. Christie and Deans showed the pathway of chloro -substituted copper phthalocyanine, up to Cl 4-CuPc using spectroscopic analysis. 37,55 It is important to mention that chloro -functionalized Pcs lead to smaller HOMO -LUMO gap and also lower LUMO. 35 The goal of this work is to establ ish the trade -off between material purity, device efficiency and material cost based on the synthesis precursors. 3.2 Experimental Procedure 3.2.1 Material Synthesis The reaction of ClAlPc from PN and PA is carried out using microwave reactor, CEM Corporation Discover SP, at 230 . Microwave -assisted synthesis of M -Pcs is not a common method, but it has advantages of reducing reaction time and increasing reaction yield. The ratio of reactants is followed by the previous study as shown in Table 3 -1.56 Additional washing is performed for PA process with sulfuric acid and sodium hydroxide solution in low concentration (0.1M). The yields of ClAlPc from PN and PA processes are 81±5% and 24 ±7% respectively. 19 Since the goal for this study is to produce low -cost OPV materials, further purification, such as thermal gradient sublimation, is not performed to evaluate if the initial ClAlPc purity is sufficient for OPV applications directly . Table 3-1. Sample description of chloroaluminum phthalocyanine synthesis from phthalonitrile and phthalic anhydride Precursor Time and Temp. Molar Ratio of reactants Solvent PN-1 phthalonitrile 40 min 230 PN:AlCl 3 4:1 1-chloronaphthalene Acetone PA-1 phthalic anhydride 40 min 230 PA:Urea :AlCl 3 4:12:1 (NH 4)2MoO4 as catalyst 1-chloronaphthalene Acetone PA-2 The s ame batch with PA -1 with additional acid/base wash 0.1M H 2SO4 0.1M NaOH 3.2.2 UV-visible Spectroscopy ClAlPc samples from PN, PA and TCI America (98 % purity for reference) are dissolved in ethanol (EtOH) through 15 min of sonication. Undissolved phthalocyanine and impurities are separated by 22 µm syringe filter. UV -vis spectra of extracts are measured through UV 2600 Spectrometer, Shimadzu, in range of 400 nm to 800 nm. 3.2.3 HPLC -MS Spectroscopy Waters Xevo G2 -XS UPLC -MS is used to analyze mass spectrum of ClAlPc samples. Mass spectra are obtained by electrospray ionization of extracts and column separation through 5% H 2O and 95 % acetonitrile as eluent. 20 3.2.4 Cost Analysis The cost of ClAlPc from PN and PA processes is estimated using the price of reactants and solvents from chemical vendors. The reference price is determined as the sales price of chemical for the largest quantity to reflect reality from Sigma Aldrich, Molbase , and TCI Americ a.57Œ59 3.2.5 Solar Cell Fabrication and Tests The device architecture is ITO (1200 Å, pre -patterned) / MoO 3 (100 Å) / ClAlPc (150 Å) / C60 (300 Å) / BCP (75 Å) / Ag (1000 Å) shown in Figure 3 -2. Layers are deposited using a growth rate of ~1 Å/s. Reference for ClAlPc is from TCI America 98%, and C 60 99.9% is from MER corporation. Device characteristic J sc, Voc, FF, and device efficiency are measured as a function of voltage (V) (Keithley 2420) under AM1.5G solar simulation (x enon arc lamp) where the intensity was measured using a NREL -calibrated Si reference cell with KG5 filter . External quantum efficiency (EQE) measurements are performed under a monochromatic incandescent halogen lamp calibrated using a Newport Si detector . Figure 3-2. Energy level diagram and structure for the solar cell based on chloroaluminum phthalocyanine and C 60 21 3.3 Result and Discussion 3.3.1 UV-visible Spectra UV-vis spectra of ClAlPc dissolved in ethanol are shown in Fig ure 3-3. UV -vis spectrum of M -Pcs is known to exhibit strong Q -band at 600 to 700 nm due to - transitions .60 ClAlPc -reference and PA solutions have maximum peak at 670 nm, but for the PN solution which has a lower pH, the peak is at 677 nm. I t is reported that acidic solutions results in red -shifts due to M -Pc protonation .60 Figure 3-3. UV-vis spectrum of chloroaluminum phthalocyanine: ClAlPc -reference (green solid), ClAlPc -phthalonitrile (blue dash -dot), and ClA lPc -phthalic anhydride (red dash) 3.3.2 HPLC -MS Spectra In Fig ure 3 -4, mass chromatograms of ClAlPc show one strong peak at 7 min of retention time with 580.16 mass charge ratio (m/z) for all samples except PA -1 samples for which there is a peak at 4 min with 487.12 m/z, suggesting significant amount of impurity. This impurity is efficiently removed by acid/base solution as shown as PA -2 in Table 1 and Fig ure 3-4. Š PA-1 22 and PA -2 are synthesized in the same batch, but PA -2 is washed with acid /base solution. All samples do not match exactly the expected m/z for ClAlPc, which is 574.96 but a slightly highe r value of 580.16. The difference could be a result of chlorine ion lost during ionization which would result in phthalocyanine adducts with acetonitrile which would correspond to the observed value. Isotope distribution is also used to verify material pur ity, and PN and PA samples match the 98% purity ClAlPc reference from TCI America. Figure 3-4. Mass chromatogram (left) and isotope distribution (right) at 580.16 m/z of chloroaluminum phthalocyanine: ClAlP c-Reference (green), ClAlPc -Phthalonitrile (blue), ClAlPc -Phthalic anhydride -1 (deep red) and Phthalic anhydride -2 (red) 3.3.3 Cost Analysis The cost of ClAlPc from PN and PA process is significantly lower than reference ClAlPc from TCI America. Although phthalic anhydride is a hundred -fold cheaper than phthalonitrile, PN 23 process produces cheaper ClAlPc since yield of PA process is much lower than PN process, 24 % compared to 81 % as shown in Figure 3 -5. Also, the price of ClAlPc is directly associated with the price of solvents, which use similar amounts for both processes. Figure 3-5. Cost analysis of chloroalum inum phthalocyanine: ClAlPc from phthalic anhydride (PA process) and phthalonitrile (PN process) compared to purchased ClAlPc from TCI America 3.3.4 Photovoltaic Performance Characteristics of ClAlPc devices from PN and PA processes are illustrated in Fig ure 3-4 and 3-5 and summarized in Table 3-2. PA process leads to the lowest device performance. The PA process requires urea addition as a source of nitrogen to form 1,3 -Diiminoisoindoline , but the urea decomposes to not only ammonia gas but also many other by products at high temperature, such as biuret, cyanuric acid, and ammelide etc. 61 Therefore, the PA process inevitably results in lower purity of ClAlPc which results in lower device efficiency. Since the devices from PN -1 sample show the highest energy conversion efficiency but yields the lowest number of working devices, we report data as both the best devices and median device performance, which show lower energy conversion efficiency although all devices work (Table 3-2). Also, external quantum efficiency 24 (EQE) of replication shows lower ClAlPc response and rare C 60 response (Fig ure. 3-7). Either impurity or degradation of ClAlPc could affect to device performance, but it is hard to diagnose the failure of device since we did not further characterize devices. Table 3-2. Photovoltaic performance parameters of optimal chloroaluminum phthalocyanine devices: ClAlPc -Reference (TCI America) , phthalonitrile (PN -1), and phthalic anhydride (PA) cells ClAlPc Source Jsc (mA/mm 2) Voc (V) FF Efficiency (%) Yield TCI America 0.05 ±0 0.78 ±0 0.50 ±0 2.1 ±0.1 8/8 PN-1 Best 0.06 0.76 0.42 2.0 - PN-1 Replication 0.03 ±0 0.06 ±0 0.42 ±0 0.80 ±0.02 8/8 PA 0.02 ±0 0.58 ±0.01 0.24 ±0 0.20 ±0.02 7/8 Figure 3-6. Representative current density -voltage characteristics of solar cells based on ClAlPc -reference (green solid), ClAlPc -phthalonitrile -1 best (blue dash -dot) and phthalonitrile -1 replication (deep blue dash) and ClAlPc -phthalic anhydride (red dot) 25 Figure 3-7. External quantum efficiency of solar cells based on ClAlPc -reference (green solid), ClAlPc -phthalonitrile -1 best (blue dash -dot) and phthalonitrile -1 replication (deep blue dash) and ClAlPc -phthal ic anhydride (red dot) 3.4 Conclusion and Future Work Two precursors of ClAlPc, phthalonitrile and phthalic anhydride, are studied to evaluate tradeoff between material purity, device efficiency, and material cost. The results show that UV -vis and mass spectra of ClAlPc from PN and PA processes are in agreement with reference material. The cost of ClAlPc from two processes are significantly lower than ClAlPc from chemica l vendor, and PN process produces three -fold cheaper ClAlPc than PA process due to high yield. Based on photovoltaic fabrication and tests, we observe two different results from ClAlPc -PN devices. The best device from PN sample shows similar energy convers ion efficiency with the reference device although fabrication yield is low. For PA samples, all devices work, but have low energy conversion efficiency due to impurities from urea decomposition. Future studies will identify trace impurities of ClAlPc from PN process and explore synthesis pathways to improve initial purity of ClAlPc. 26 Chapter 4 Fine Chemical Process Toward to Sustainable Organic Photovoltaic The p revious chapter s discu ss the experimental and simulation method to synthesize ClAlPc and to fabricate ClA lPc solar cell, and to assess life cycle assessment of OPV. Those methods are applied in this chapter to assess fine chemical process to evaluate environmental, cost, and chemical hazard impacts and demonstrating iterative design approach . As described in the motivation, establishing a holistic evaluation method for OPV material process is important to develop sustainable energy solution. This work has been published as E. Lee, C. Andrews, and A. Anctil, fi An Iterative Approach to Evaluate and Guide Fine Chemical Processes: An Example from Chloroaluminum Phthalocyanine for Photovoltaic Applications fl, ACS Sustainable Chem. Eng., 2018, 6 (7), pp 8230 Œ8237. Additional information regarding data source for life cycle assessment, synthe sis process, material purity, and evaluation criteria for ClAlPc processes is available in the appendix . 4.1 Introduction Fine chemicals are complex , single, and high purity chemical substances produced in limited quantities (up to 1000 MT per year) at a high price (>$ 10/kg) for use in the pharmaceutical, agrochemical, animal health, pigment, and electronic industries. 62 One method that can be used to reduce the envir onmental impact of fine chemical manufacturing is green chemistry, which is well -known for its 12 principles. 13 They are general guidelines to reduce the environmental impact of chemical synthesis, for example, by reducing byproduct and derivatives, increasing energy efficiency, and encouraging safer chemical use. The framework of green chemistry requires (1) consider ing all stages of the chemical life cycle, (2) try ing to reduce the intrinsic hazard of chemical products by design, and (3) us ing green chemistry as a cohesive system of principles or 27 design criteria. 13 Anot her method to reduce the impact of a product is life cycle assessment (LCA) which is a comprehensive method for assessing impacts associated with the whole life cycle of a product, from material extraction (cradle) to final disposition (grave). LCA quality is dependent on the accuracy of mass and energy flow information collected to construct the life cycle inventory. To remain competitive , the fine chemical industry rarely publishes detailed information through patents or journal article publications. Ener gy consumption directly associated with one chemical is also difficult to assess since fine chemicals are produced in multipurpose plants where equipment and facilities are shared between production lines and are used for multiple chemicals. 63 In addition to academia, pharmaceutical and chemical companies have established various sustainability methods that combine aspects of LCA and green chemistry to guide their product development. 64Œ66 A total of 19 methods is summarized and compared to the proposed methodology in Table A1 of appendix . Existing methods for industrial a pplications require extensive data collection to either perform a full LCA or calculate company specific impacts. At the other end of the spectrum, lab scale methods tend to use simplified impact assessment focusing on the manufacturing stage of the proces s, without consideration of upstream chemicals used as LCA would do. To the knowledge, there is no screening method to evaluate the impact of manufacturing and upstream chemicals on the basis of environmental impact, cost, and chemical hazard criteria that can be used to guide chemical development at the lab scale. LCA is a useful tool to evaluate chemical and pharmaceutical processes as long as the inventory of all chemicals and processes already exists, which is not the case for most new chemicals. Also, the characterization factors which are necessary to calculate the environmental and health impacts are also often missing. Without characterization factors for certain chemicals in the LCA database, the assessment will underestimate the impact. Green chemi stry principles 28 can help identify issues associated with new chemicals regarding toxicity and material efficiency and can be used to complete the information from the life cycle assessment. This work propose s a new evaluation method to rapidly screen poten tial environmental, health, and cost impacts for fine chemicals with limited LCA information that can be used by both industry and academia. The goal of this study is to demonstrate the use of an iterative sustainability methodology that combines process -based LCA and green chemistry metrics to evaluate and guide fine chemical synthesis. The proposed method is applied to small molecules which are used in organic photovoltaics (OPV). 67 Current OPV research focuses on reducing cost and increasing the efficiency of solar cells, and environmental and health impacts from OPV materials are mostly unknown. 68,69 The basic structure of OPV requires mixing a donor and an acceptor material, 70 such as metal phthalocyanine (M -Pc) as a donor and fullerenes as an electron acceptor. 71,72 Chloroaluminum phthalocyanine (ClAlPc) is of particular interest for OPV since it absorbs only in the near -infrared region. It allows the solar cell to be transparent, which is important for applications such as power -generating windows. 11,30 While metal phthalocyanines are common small molecules widely used in the dye industry, 33 the required high purity for s olar applications 32 makes it a fine chemical. Current prices for high purity (98%) ClAlPc is $157/g 73 compared to $55/g for lower purity (85%). 74 The price difference can be attributed to the energy intensive purification process necessary for electronic applications. 12 The proposed methodology is iterative and follows the steps shown in Figure 4-1(a). It begins with the synthesis of ClAlPc which will be considered the baseline, using the conditions reported in a published patent. 38 During the synthesis, the direct energy, chemical consumption, and yield are measured to build the life cycle inventory. The material needs to meet the purity requirement for OPV applications and is characterized using UV -Vis and HPLC mass spectroscopy. Impact 29 assessment which uses a combination of green chemistry impact factors and LCA is performed using the life cycle inventory. The process is evaluated based on the three impact categories shown in Figure 4-1(b). The fihotspotfl is identified as the chemical or process which has the highest contribution for one or more of the impact categories. An alternative is developed using green chemistry principles and compared to the baseline. The iterative method is repeated until an alternative that simultaneously reduces the environmental, cost and chemical hazard is found . As sho wn in Figure 4-1(b), the goal is to design for sustainability and therefore the alternatives need to reduce all impact categories and avoid trade -offs. For example, designing for lower cost, could have the unintended consequence of increasing the chemical hazard and environmental impacts. Figure 4-1. (a) Proposed iterative methodology and (b) sustainability method and categories 4.2 Method 4.2.1 Life Cycle Assessment for Fine Chemicals Process The study includes all stages from raw material extraction to the production of 1 kg of ClAlPc to meet purity requirement of OPV application. The use of LCA software facilitates the compilation and analysis of the inventory data and for this project, SimaP ro (Pre Consultants, 30 Netherlands), is used , in combination with inventory data from existing databases (Ecoinvent 3.3 and US -EI from DATASMART) 75,76 previously published literature, and data collected during the chemical synthesis. Inventory data for new chemicals is estimated using default values and stoichiometric reactio ns based on previously published guidelines. 63,77,78 The standard transportati on distance for chemicals is adapted to reflect the U.S conditions using the average mileage of commodity flow survey. 79 Additional details on inventory assumptions and life cycle assessment are available in Table A2. The goal is to combine life cycle assessment method for existing chemicals with green chemistry metrics to assess the impact of material synthesis and rapidly screen potential concerns. All 12 green chemistry principles are considered to be included since they were not meant to be independent goals. 13 Care is taken to not duplicate information from the life cycle assessment with green chemistry metrics but rather complement it . Sustainable chemical process design requires reducing the en vironmental, economic and social impacts simultaneously. Since the objective is to provide fast screening of chemical process only sub -categories that can be assessed at this stage of product development are used . Table 4-1 summarizes the metrics and sub -criteria selected in the methodology. Environmental impacts are commonly assessed using LCA, and therefore, for existing chemicals, those sub -categories are chosen and complemented with green chemistry metrics. For economics assessment, the production rate provides an estimate of maintenance and facility use in addition to the life cycle cost of raw materials. Finally, for the social impact, the only aspect that can be characterized and minimized during R&D is the intrinsic hazards of the chemicals to worker s. Therefore, chemical hazard is used as our third impact category although both exposure and toxicity assessments of chemicals should be considered for risk characterization. 31 In Table 4-1, only two principles are used as sub -criteria because LCA methods are chosen instead. Four principles are not applicable, including #4: Designing safer chemicals and #10: Design for degradation since we do not want to change the chemical. The other five principles such as # 5: Safer solvents and auxiliaries are used to g uide alternative synthesis rather than in the evaluation stage (Table 4-1). Table 4-1. Summary of green chemistry metrics and LCA impact categories used in the assessment methodology and to guide alternative synthesis Green Chemistry Life cycle assessment Principle (#) Metrics or tool Impact category Method Sustainability Methodology Index Environmental E4: Waste prevention (1) E-Factor: amount of waste generated /kg of product E1: global warming potential (GWP) Tool for reduction and assessment of chemicals and other environmental impacts (TRACI) 2.1 - U.S. 2008 normalization 80 E2: Cumulativ e energy demand (CED) CED 1.09 81 E3: water demand indicator (WDI) Water accounting and vulnerability evaluation (WAVE) 82 Chemical hazard H1-H4: Inherently safer chemistry for accident prevention (12) NFPA 704 standard (0 -4 score): H1: health hazard H2: flammability hazard H3: reactivity hazard H4: special hazard Cost C1: Life cycle cost Price of petroleum, coal, and natural gas as raw material inputs C2: Production rate Reaction capacity based on vessel size and time per batch Guide Alternative Synthesis Atom Economy (2) Atom economy: formula weight (FW) of atoms utilized/ FW of all reactants in percentage Less Hazardous chemical synthesis (3) Globally Harmonized System 83 Safer Solvents and Auxiliaries (5) GSK Solvent Selection Guide 84 Design for Energy Efficiency (6) Ambient temperature and pressure process Reduce derivatives (8) Atom economy: FW of atoms utilized/ FW of all reactants in percentage Catalysis (9) M-Pc synthesis with organic base 85 Not applicable Designing safer Chemicals (4) Preferred molecular design to reduce chemical toxicity Use of renewable feedstock (7) Raw material or feedstock should be renewable Design for degradation (10) Designing Small Molecules for Biodegradability 86 Real -time analysis for pollution prevention (11) Real -time, in -process monitoring and control prior to the formation of hazardous substances 32 Weighting and normalization are used to convert each sub -criterion to one dimensionless impact category unit and, for this work, all sub -criteria are given equal weight. The baseline process is given a 100% score for each category to allow relative compari son of the alternatives. 4.2.2 Selection of Impact Category for Sustainability Assessment 4.2.2 1 Environmental Impact Global warming potential (GWP ) (E1) and Cumulative energy demand (CED) (E 2) are common environmental impact indic ators with a well -established methodology. If inventory data is not available, the new materials inventory guidelines mentioned before are used but those methods are limited to CED and GWP. For water use, the water demand indicator (WDI ) (E3) is a LCA metric that includes water consumption from chemical process and utility input which is not accounted for in the E -factor. Although E -factor have recently included water use 87, the evaluation method follows the original definition of th e E-factor and does not. The green chemistry metrics listed in Table 4-1 are calculated in Table A6 and used to develop alternative ClAlPc process. The first of the green chemistry principles, #1: waste reduction , is key to reducing the environmental impact of chemical process. It is calculated using the E -Factor (E 4) which is the mass balance based on actual process of wasted material. 2,88 Similarly, process mass intensity (PMI ) is defined as the ratio between quantity of raw materials involved in the process and quantity of product in kg. PMI includes water use and therefore E-factor is used as an impact category instead since water consumption is already calculated from the WD I.2,35 33 4.2.2 2 Chemical Hazard Chemical hazard is used to estimate the social impact. According to the World Business Council For Sustainable Development (WBCSD), workers™ occupational health risks and safety are considered a mandatory topics in social life cycle metrics for chemical products. 89 The recommended method for LCA human health impact is the consensus model USEtox developed by UNEP/SETAC and based on the on the potential toxicity impacts and emission of chemicals. 90 This method cannot be used since ClAlPc processes include uncharacterized chemicals with unknown toxicity and emissions. EPI suite and ECOSAR use quantitative structure -activity relationship models (Q SAR ) to respectively estimate the environmental fate and aquatic ecotoxicity of chemicals . Their use is limited to organic chemicals since chemical descriptors are not suitable for inorganic molecules and organometallics such as aluminum chloride and ClAlPc used in our alternative processes. 91,92 For these reasons, only green chemistry metrics are considered for chemical hazard: #3: Less hazardous chemical synthesis , #5: Safer solvents and auxiliaries , and #12: Inherently safer chemistry for accident prevention. Principles 5 and 3 are used to guide alternative synthesis, and principle 12 is used to ev aluate chemical hazard (Table 4-1). H 1-H4 are based on the NFPA 704: standard system for the identification of the hazards of materials for emergency response . NFPA 704 standard is based on acute exposure to chemicals, such as fire, spill or similar emergency. It provides information regarding the health, flammability reactivity, and special hazards of materials that are assessed by short -term and acute exposure and use a rating system fr om 0 to 4, where 0 is low hazard. 38 The Globally Harmonized System (GHS ) is also a classification and labeling system but it covers additional aspects such as chronic toxicity and environmental hazard. However, an ove rall score is not provided even though it can be done as demonstrated by the GSK solvent guide 84 where a decision tree was used to assess health hazard. 34 Based on the decision tree methodology, the health hazard score is determined by occupational exposure limits (OELs) or GHS Hazard and Precautionary risk phrases. If a chemical does not have an OEL value and is annual production is less than 1,000 tons/yr, the health hazard score is 4 out of 10. In the ClAlPc case, most health hazard score for chemicals used in the processes would receive this score since their annual production is lower than the 1,000 tons/yr. Therefore, the NFPA 704 standard is selected t o reflect actual health hazard. The hazards scores for each chemical are multiplied by the mass of chemical used to produce 1kg of ClAlPc and then normalized based on the baseline score. Health (H 1) and flammability (H 2) hazards are each weighted as each 1/3, but reactivity (H 3) and special (H4) hazards are equally weighed 1/6 respectively since the oxidizing property of chemicals is responsible for both the reactivity and the special hazards scores. 4.2.2 3 Cost Cost is characterized by C 1, the life cycle co st and C 2 the production rate (Table A7 and A8) There are no green chemistry metrics that directly assess the cost of the process although some have used PMI as an indicator. 24 PMI is based on the mass of chemicals, including water, not the specific chemical price and therefore the use of expensive chemicals, in particular for fine chemicals is not well reflected by the metric. 4.2.3 Material Synthesis and Data Collection The baseline process for ClAlPc is based on a patent 38 and uses a heating mantle. Alternatives processes are numbered as P1 -P5 which correspond to the order in which they are developed (Table 4-2). The reaction of ClAlPc from phthalonitrile (PN) and phthalic anhydride (PA) is 35 carried out using either heating mantle or a microwave reactor (CEM Corporation Discover SP). In the M -Pc synthesis, the molar ratio of PN or PA to metal ions is fixe d to 4:1. 33,36,37 The source of metal ion is aluminum chloride (AlCl 3), and PN or PA are mixed with the metal salt in reaction media. High boiling point solvents are needed for all sy nthesis because of the high reaction temperature (230 C). For microwave synthesis, polar solvents are required because heating is based on molecular rotation. 47 1-chloronaphthalene (1-ClNP ), 2,4 -dichloroanisole (2,4 -DCA), and diethylene glycol (DEG ) are used as reaction media. Auxiliaries such as urea, ammonium chloride (NH 4Cl), and 1,8 -Diazabicyclo[5.4.0]undec -7-ene (DBU ) are added for P2 -P5. Table A2 provides details about the synthesis conditions. Direct energy demands, such as electricity from the reactor and vacuum pump, and chemical consumption are recorded during the synthesis of ClAlPc. To better reflect industrial process , the theoretical energy deman d for the baseline process is calculated using previously reported method 77 and converted to steam. For material characterization, ClAlPc samples and reference (TCI America 98 %) are dissolved in ethanol through 15 min of sonication. Undissolved phthalocyanine and impurities are removed using a 22 µm syringe filter. 60 UV-vis spectrum of extracts is measured using UV -Vis 2600 Spectrometer, Shimadzu. Waters Xevo G2 -XS UPLC -MS is used to analyze the mass spectrum of ClAlPc samples. Mass spectra are obtained by electrospray ionization of extracts and column separation with 5% H 2O and 95 % acetonitrile as eluent. Additional details on material characterization are available in Table A 8-S9. 36 Table 4-2. Reaction co ndition and chemical use for ClAlPc processes Process Energy source Precursor Reaction media Auxiliaries Rxn. Time Type Use (g) Type Use (g) Type Use (g) Baseline Heating mantle PN : AlCl 3 1.80 : 0.47 1-ClNP 10.75 . . 6 hr 1 Microwave PN : AlCl 3 1.80 : 0.47 1-ClNP 10.75 . . 40 min 2 Microwave PA : AlCl 3 1.80 : 0.41 1-ClNP 5.44 Urea 2.19 40 min 3 Microwave PN : AlCl 3 1.80 : 0.47 2,4-DCA 6.79 Urea 0.54 40 min 4 Microwave PN : AlCl 3 1.80 : 0.47 2,4-DCA 6.79 NH4Cl 0.54 40 min 5 Microwave PN : AlCl 3 1.80 : 0.47 DEG 6.66 DBU 2.48 40 min 4.3 Result and Discussion 4.3.1 Methodology Assessment The iterative method is shown in Figure 4-1 (a) is used for ClAlPc synthesis and alternative processes developed are summarized in Table 4-1. Figure 4-2 (a) illustrates the iterative process and the identified hotspots for the baseline, P1, P4, and P5 as examples. Figure 4-2 (b) -(e) shows the impact assessment for each p roposed alternative compared to either the baseline or the previous best chemical synthesis process. The impact for each category is fixed at 100% which correspond to the baseline score to facilitate comparison between the proposed alternatives. Summary of the score for each process is available in Table A 9. The following sections summarize the alternative processes from the baseline to P5. Baseline to Process 1. The baseline process is a well -established pathway for ClAlPc synthesis using solution processing and has the highest yield at 84%, but high yield is not necessarily synonymous with a sustainable process. The two main concerns or fihotspotsfl from the baseli ne process are the reaction time ( six hours) which impacts production rate as well as the toxicity of the reaction media. The purity requirements for OPV application are also not met due to partial chlorination of ClAlPc, which can be considered an impurit y (Figure A9). 37 To address these issues , the use of microwave rather than heating mantle and alternative solvents is considered, based on reported conditions for other M -Pc. 44,93,94 P1 uses microwave reaction rather than heating mantle but keeps all the other conditions as the baseline. The reaction time is reduced from six hours to 40 min which is an improvement compared to the baseline, but longer than the five to ten minutes repor ted for other M -Pcs. 44,93 Figure 4-2. (a) Summary of hotspots determined by iterative evaluation , and (b) -(e) evaluation of alternative ClAlPc processes based on environmental, chemical hazard and cost impacts 38 The microwave reaction increases both the environmental and the cost impact as shown in Figure 4-2(b). The microwave process is faster, which increases the production rate by 27 %even when considering a smaller reaction volume (3.5 L for microwave reactor versus 20 L of pilot scale reactor). However , the electricity consumption in the microwave process is higher and the chemical yield lowe r (73% vs 84%) compared to the baseline which results in a 31% increase in environmental impact. The cost impact does not change significantly. Baseline to Process 2. Since P1 increases the environmental impact, but result in no reduction in cost, the se cond alternative pathway (P2) considers alternative precursors for ClAlPc. PA is used by the dye industry to synthesize CuPc because of its low cost. 33 The bulk chemical of PA is three times cheaper than PN74 and therefore could significantly lower ClAlPc price. In the Porphyrin Handbook , various ClAlPc processes are summarized including reaction condition and yield, but none of process uses PA as a precursor. 36 Unfortunately, P2 reaction yield is only 25% (Table A5) and the impact for all 3 categories increases. Previous work by our group has shown that ClAlPc synthesis from PA produces impurities due to urea decomposition that reduces photovoltaics device efficiency (Figure A8). 95 Using PA does not seem like a suitable pathway for ClAlPc synthesis since it reduces material purity while increasing enviro nmental and cost impacts. Therefore, P2 is excluded from Figure 4-2 and the next iterative step is based on P1. Process 1 to Process 3. The initial process modification P1 tried to address one of the two hotspots from the baseline process, the long reactio n time but the use of microwave resulted in an increased electricity consumption. The second hotspot which is the toxicity of 1 -ClNP is considered for P3 using the green chemistry principle #3: Less Hazardous Chemical Synthesis . Chlorinated naphthalene is known to be toxic to aquatic environment and has been prohibited for industrial use in Japan due to its persistency in aquatic environment. 96,97 The fifth green 39 chemistry principle: safer solvents and auxiliaries is used to find alternative solvents using solvent selection guides. 84,98 Replacement of 1 -ClNP is difficult due to the high reaction temperature. Based on the CHEM21 solvent selection guide, anisole would be a good alternative solvent but has a boiling point of 155.5 C. It has been reported that using anisole instead of a phenoxy derivatives for ClAlPc synthesis produces a high crystallinity compound. 99 We select 2,4 -dichloroanisole (2,4 -DCA) as an anisole alternative since it has a boiling point of 285 C. For the baseline and P1 synthesis, extra volume of solvent was used to prevent chlorination of phthalocyanine, but it limits the reaction volume per batch and therefore the production rate. To prevent partial chlorination of ClAlPc , 30 w/w % of urea is added while reducing the volume of 2,4 -DCA. 33,37,55 P3 compared to P1 reduces the chemical hazard by 23 % but both environmental and cost impacts increase (893% and 352% respectively). The increase in environmental impact is associated with high CED and GWP due to the upstream process of 2,4 -DCA. The chlorination of anisole requires large volumes of organic solvents and acetic acid, which is responsible for 40% of the CED and GWP impacts of 2,4 -DCA. The importance of LCA in our method is demonstrated for P3 sinc e it provides insight on the impact from the upstream stages which is not considered by green chemistry alone. Another issue from the 2,4 -DCA synthesis is filtration. The crude ClAlPc solution becomes viscous after the reaction is completed and the separat ion requires a larger volume of organic and inorganic solvents compared to the baseline method which increases the E -factor and water demand. In term of cost, P3 raw material cost is 637 % higher than the baseline due to the 2,4 -DCA cost and the extra amount of washing solvents. Overall the cost of P3 is 352% that of the baseline 40 due to a higher prod uction rate and lower use of reaction media which counterbalance the higher raw material cost. The material characterization using HPLC -MS shows the same impurity observed in P2 (Figure A8). Therefore, P4 focuses on another source of ammonium to solve the purity issue. Process 3 to Process 4. The use of urea as a reacti on auxiliary in P2 and P3 results in insoluble byproducts which cannot be filtered out due to the decomposition of urea between 200 -250 C.61 The alternative source of ammonium for P4 is ammonium chloride due to its high boiling point (338 C). The salt is stable and soluble in water after reaction which allows its removal during filtration. P4 solves the impurity production from urea decomposition, in addition , to increase the yield by 10% compared to P3. The environmental and cost impacts of P4 are lower than for P3 but remains much higher than the basel ine (733 % and 284 % respectively) as shown in Figure 4-2(c). So far, the transition from baseline to P4 results in a reduction in chemical hazard but increases the environmental and cost impacts. Therefore, the approach demonstrates the potential of creat ing new issues while developing alternative synthesis if we focus on a single issue. The concern with the toxic solvent has not been solved since 2,4 -DCA used in P3, and P4 shifts the problem from chemical hazard to environmental and cost impacts. The next step is therefore to reduce the environmental and cost impacts which are associated with 2,4 -DCA synthesis. Process 4 to Process 5. To find an alternative to 2,4 -DCA, GSK™s green solvent list is used . Both ethylene glycol and diethylene glycol (DEG) are listed on the list and have the same score but the boiling point of ethylene glycol is lower than the required reaction temperature. 84 DEG is chosen DEG as reaction media and DBU as a proton acceptor for P5, corresponding to #9: Catalysis . 41 The CED and GWP associated with the upstream processes of DEG and DBU is smaller than 1-ClNP although direct energy use from microwave is higher. In P5, a single washing solvent, methanol which is a safer solvent (principle #5), reduces the overall environmental impact by reducing the e -factor and WDI. DEG is a co -product of ethylene glycol, which is produced in large quantity globally 100 and therefore is cheaper than 2,4 -DCA and 1 -ClNP. Using chemical produced in large quantity has the advantage that toxicity of the chemical has been tested since it is required for all chemicals with a production higher than 1,000 tons per year. 101 The use of DEG as reaction media reduces the environmental and cost impacts compared to P4 while chem ical hazard remains similar as shown in Figure 4-2(d). 4.3.2 Overall Evaluation Using this iterative method , the baseline process is modified to P5 which reduces the environmental, cost and chemical hazard impacts as shown in Figure 4-2(e). The identification of the hotspot for each process effectively guided alternative processes. In P2, atom economy might be used as a screening criterion for chemical reaction selection but used alone it can lead to synthesis methods with higher environmental impact due to lower yield which results in additional material and energy use from the upstream process. For microwave reactor, microwave is known to reduce reaction time for phthalocyanine synthesis 44,102,103 , but scaling up is challenging due to the high power consumption. Although industrial scaling up for microwave reactor is not considered in our study, there is a commercial microwave plant in Japan which suggest that microwave could be used at an industrial scale to produce ClAlPc. 104,105 The location is important for CED and GWP since the main contributor to these impact categories is electricity which dep ends on the portion of fossil fuel in the electricity mix. 106 42 Chemical production is energy intensive and energy -related environmental impact can account for almost half and sometimes up to 80% of GWP. 107 Figure 4-3 compares the CED impact from some of the processes considered. For P1 and P5, the electricity consumption from the microwave h as the highest impact when considering the chemical synthesis and separation. For P4, as previously discussed, the solvent upstream process is more important than the direct energy consumption. Figure 4-3. Cumulative energy demand for baseline, 1 -chloronaphthalene (P1), 2,4 -dichloroanisole (P4), and diethylene glycol (P5) ClAlPc processes In addition to the direct electricity consumption, the relative contribution of other fa ctors to CED is illustrated in Figure 4-4 for the baseline and P5 for all stages from basic chemical, synthesis, and separation. The direct electricity consumption (orange in Figure 4-3) and transportation is excluded from the figure to emphasize the impac t of chemicals. In Figure 4-4, the upstream process of 1 -ClNP has higher CED impact compared to the combination of DEG and DBU (865 MJ compared to 560MJ). 43 (a) (b) Figure 4-4. Sankey diagram of the cumulative energy flow for ClAlPc synthesis from (a) the baseline and (b) diethylene glycol and DBU (P5) excluding transport and electricity This work developed of a new methodology that combine LCA and green chemistry metrics to eval uate and guide fine chemical manufacturing. A secondary contribution is the creation of life cycle inventory for the four processes developed for ClAlPc based on a microwave reactor. Finally, a new method P5 is developed to produce ClAlPc which reduces sim ultaneously the environmental impact by 3%, the cost by 9% and chemicals hazard by 23% compared to the baseline process while producing high quality material that can be used for organic photovoltaics application. 44 Chapter 5 Building Energy Model for Life Cycle Ass essment of Transparent Organic Photovoltaic in Window Application This work includes a simulation of the building energy demand by TPV in window application. The material of TPV used in this chapter is ClAlPc which is discussed in the previous chapter s. 5.1 Introduction The energy consumption from fossil fuel is becoming an important issue due to resource depletion. In the United States, fossil fuels are the largest source of energy for electricity generation by about 63%.108 The primary e nergy consumption from the residential and commercial building sectors is about 50% .109 An approach to solve this large -scale en ergy demand in the building sector could be building -integrated photovoltaics (BIPV) application. BIPV are integrated directly into facades or other surface s of a building so that they replace conventional building materials and save the energy used in a building . A previous study suggests the use of Transparent PV (TPV) as a window for future energy solution since TPV can optimize both average visible transmission (AVT) and power conversion efficiency (PCE) , thereby providing sufficient energy for the U.S. electricity consumption. 11 The same study also estimates 5 to 7 billion square meter of the glass surface in the U.S., and there will be an additional 100 GW if TPV with 15% PCE were applied as window and skylight to the glass surface .11 Using TPV in window application could manage energy demand of building by absorption of near -inf rared radiation , which is felt as heat . The function of heat management by semi -transparent PV has been studied and has resulted in saving energy from air conditioning. 110Œ113 According to the Commercial Buildings Energy Consumption Surve y (CBECS ) from the U.S. Energy Information Administration (EIA ), electricity use is higher 45 in the summer due to the cooling and ventilation while the use of natural gas is mainly for heating during winter. 114 TPV in window application can mitigate these pattern s of energy use since electricity generation in summer will be higher. Natural gas use could also be saved by internal reflection of longer wavelength radiation related to heat energy in a particular area with a mild climate. The use of semi -transparent PV for heat management has been evaluated based on annual electricity cost 18 and building energy demand. 18,111,112,115 According to previous studies, semi -transparent PV application helps to reduce cooling energy demand in a warm climate but increase s heating energy demand in cold climate because less solar energy radiates into the building . Most studies assess silicon -base d solar cell such as amorphous silicon PV 18,111,112,115 , micromorph silicon PV 111, and only one paper consider s organic PV 112 with limited information rega rding optical properties (60% absorption and 30 % transmittance) and PV efficiency (3% PCE) . All PV studied in previous studies are semi -transparent PV that have low visible transmittance of around 10 to 20%. There is only one experimental study that demonstrates the benefit of semi -transparent PV115, whereas other studies use simulation software of building energy de mand, such as EnergyPlus managed by the National Renewable Energy Laboratory (NREL). A study consider s the thermal balance of PV application in a building energy system that accounts for solar radiation transformed into electricity 112 , while others use PV optical properties such as window properties in th e simulation. All papers do not consider the lifetime of the window and PV so that analysis is based on annual energy demand. Similar to the window application of PV, there is an LCA study applying PV as adaptive shading. 116 The concept of the PV applica tion is not as a window , but as the shading application , and it shows that such application reduces cooling and heating energy 46 demand. The LCA is based on 20 year s™ application of adaptive shading by PV application , and the system boundary includes PV, balance of systems (BOS), building operation, and disposal. The study assesses G WP, and other six major ReCiPe midpoint indicators (TAP, FEP, HTP , MDP, and POFP). This chapter discusses the energy demand of commercial building -applied TPV and reports the building energy simulation data of medium office from reference building structure developed by U.S. Department of Energy (DOE) by using EnergyPlus . The results of building energy demand will be used as background data for asse ssing net energy benefit (NEB ) of TPV applica tion for 20 years ™ lifetime through LCA. 5.2 Method The goal of this task is to assess building energy demand for evaluating net energy benefit (NEB) of transparent organic photovoltaic s (TPV) in window application based on the simulation of building energy demand. The o verall framework of t his study includes material production, module manufacturing, and module use as illustrated in Figu re 5-1(a) , but the contribution of this chapter is limited to the use of TPV module as a window . TPV are fabricated on glass substrates that are pre -coated with 120 nm thick of indium -tin-oxide (ITO) as a function of the transparent anode. Molybdenum(VI) oxide (MoO 3, 10 nm), chloroaluminum phthalocyanine (ClAlPc, 15 nm), fullerene (C 60, 30 nm), and bathocuproine (BCP , 7.5 nm) are sequentially deposited through thermal evaporation . Then indium tin oxide (ITO , 100 nm) cathode is rf -sputtered directly onto the organic layer. 30 Since the ClAlPc -based device has high -visible transparency of > 55%, it could substitute the original function of window s. The busbar and silver grid are assumed to be applied in the system for expanding laboratory scale 47 data of TPV to real application. Total coverage of the busbar and grid is 11% of the window area with 3 nm thick titanium and 70 nm thick silver based on previous literature of transparent electrode s.117 Figure 5-1. System diagram shows : (a) scope of simulations in building system simulation and photovoltaic simulation for TPV application in window , (b) system boundaries of l ife cycle assessment including material production and module manufacturing , and module use of TPV as windows and skylights , (c) scenarios of TPV application . baseline window composites of 3mm glass and 13 mm air gap; glass TPV that contains the active layer of TPV represents window replacement; Plastic TPV encapsulated with PET film is a scenario of retrofitting window . Clear double pane window is used for the baseline simulation for window or skylight structure 18, and Figure 5-1(b ) illustrates the baseline and the two alternatives considered for TPV , the first case with TPV located inside the double pane window and the second case with TPV encapsulated with PET film. Since TPV is based on organic material, it could degrade under 48 oxygen or moisture condition s.118 Therefore, TPV is either deposited inside of the two panes of the window (glass TPV) or encapsulated with the film (plastic TPV) as shown in Figure 5-1(c). 5.2.1 Selection of study area The selected cities are Detroit , MI, Los Angeles, CA, Phoenix, AZ, and Honolulu, HI , since they represent a range of solar insolation conditions, climate zones, energy cost and environmental impact of electricity production . These factors will influence th e benefit of both PV electricity production and building energy balance. Figure 5-2 shows the climate conditions of the cites selected as study area s. Figure 5-2. Annual climate conditions (global horizontal irradiance, relative humidity, and dry bulb temperature) of study areas 49 Global horizontal irradiance (GHI) is the total irradiance from the sun that is the sum of direct irradiance and diffuse horizontal irradiance. Honolulu has the highest GHI, followed by Pheonix, then Los Angeles, and Detroit. The dry bulb temperature varies more in Detroit and Pheonix while it does not vary much in the other locations. The relative humidity is relativ ely higher in the three cities of Detroit, Los Angeles and Honolulu , while Pheonix is dry, especially in summer. Detroit, MI Los Angeles, CA Phoenix, AZ Honolulu, HI Latitude 42.33 33.90 33.44 21.31 Grid eGrid region 106 RFCM CAMX AZNM HIOA Climate zone IECC 2015 119 5A 3C 2B 1A Thermal Criteria 120 3000 < HDD 18°C 4000 HDD 18°C 2000 3500 < CDD 10°C 5000 5000 < CDD 10°C Insolation Fixed tilt(deg.) 0/90, (Annual average, kWh/m 2/day) 121 3.8/2.9 4.9/3.5 5.7/4.0 5.4/2.9 Figure 5-3. Summary of study area s based on the electricity grid and climate zone . eGrid is a source of data on the environmental characteristics of electric power generated in the U.S. Thermal criteria classify IECC climate zone based on degree days. Figure 5-3 summarizes climate zone , emissions & Generation Resource Integrated Database (eGrid ) sub -regions in 201 6 106, and solar insolation .121 The climate zone from 1 to 8 is 50 determined by temperatu re variations represented as heating degree days (HDD) and cooling degree days (CDD) 120, but zone 8 is excluded in Figure 5 -3 since it is only for Alaska which is not included in the study area selection. The degree days are a measurement of the demand for energy needed to heat or cool a building. 122 The climate zones are then divided into moist (A), dry (B), and marine (C) subregions based on mean temperature and precipitation. 120 5.2.2 Simulation for Building Energy Demand and Ele ctricity Production EnergyPlus and the DOE reference building s are used to simulate building energy demand . EnergyPlus is an energy ana lysis and thermal load simulation program and calcu lates the heating and cooling loads to maintain the setpoints of thermal control through heating, ventilation, air conditioning (HVAC) system. 123 It calcu lates the energy demand of the building s from the heat and mass balance and building system simulation s. The rmal properties of building exteriors, such as envelopes and windows (Table 5 -2), are used to calculate the heat and mass balance, and the energy demand of building based on electricity and natural gas is determined based on the specified size and type of HVAC system (Table 5 -1).124 Another simulation softw are used to calculate electricity production from TPV in window application is RETScreen. RETScreen is a clean energy management software for energy efficiency and renewable energy . The performance of a photovoltaic system is influenced by design elements , such as the amount of solar radiation on PV, the type and area of PV, its power conversion efficiency , and its slope and azimuth (physical orientation). The power conversion effici ency of TPV is assumed to be 10% as considered by the previous study, 11 and the other details of electricity production simulation s, such as the slope of skylight and PV areas, are summarized in Table 5-1. 51 The simulation method for building energy demand follow s previous work and uses default values for building internal loading .18 Building geometry and heating source s are mod ified from the DOE reference building of post -1980 construction 124 through SkechUp Pro 2017 to obtain the desired window to wall ratio (WWR ) or skylight to roof ratio (SRR ) (Table 5-1). The thermal chara cteristics of building envelop are listed in Table 5-2. The glazing system is assumed to be double -pane glazing system as shown in Figure 5-1(c). Table 5-1. Simulation inputs for building energy demand and TPV application Medium office Small office Building geometry Building orientation: south Building orientation: south PV application Window Skylight HVAC System type 124 Variable Air Volume (VAV) with re -heat terminal Packaged single -zone (PSZ) Heat 124 Natural gas heating and electric heating coil Cooling 124 Direct expansion cooling coil Window to wall ratio or skylight to roof ratio (%) 67.8 14.9 Window area (m 2) 938.2 (SEW) / 1340.3 (SEWN) 81.0 (S) PV area (m 2) S1: OPV in window S2: OPV in PET film S1: OPV in window S2: OPV in PET film SEW: 835.1 SEWN: 1193.0 SEW: 799.0 SEWN: 1141.4 72.1 (S) 68.9 (S) Subarea Subarea S and N: 357.9 E and W: 238.6 S and N: 342.4 E and W: 228.3 The mid -size office reference building is modified to a maximized WWR and used to study window application. TPVs are applied to windows in all directions . Since TPV application to the 52 north could result in lack of solar radiation passing through building inside and thus increase overall energy demand of the building , a simulation case without the north window (SEW versus SEWN) is considered . The small -size office for skylight application is also modified to incorporate a skylight onto the south roof and attic space is integrated to the office area as shown Table 5-1 (building geometry). In the EnergyPlus simulation, the azimuth orientation for both types of building is assumed to be south, and the weather data for study locations is importe d from EnergyPlus website to obtain climate data and schedule of building operation . 125 Table 5-2. Thermal properties of envelope used for DOE reference building in each study area Study Area Climate Zone Envelop thermal transmittance (U-factor, W/m 2.K) Roof Wall Medium office 124 Small office Medium office 124 Small office 124 Insulation entirely above deck Attic and other Steel frame Wall Mass Wall Detroit 5A 0.30 0.31 0.47 0.57 Los Angeles 3C 0.57 0.63 1.25 5.68 Phoenix 2B 0.26 0.27 1.36 2.33 Honolulu 1A 0.42 0.46 5.68 5.68 In EnergyPlus simulation, three spectral properties of a window which are transmittance, front reflectance, and back reflectance, are required as input parameters to calculate the building energy demand . The TPV in window application transmits visible radiation and absorbs near -infrared radiation to generate electricity. The electricity generation by TPV in window application is not currently supported in EnergyPlus, and the absorbed radiation is considered as heat and emitted indoors. Therefore, the absorbed radiation that could generate electricity causes an extra heat flowing into the window and thus results in an overestimation of required cooling e nergy. To solve this issue , a previous study suggested adding an additional reflectance on the front surface of the outside glass followed by energy conversion efficiency of solar cell .112 Thus , an additional reflectance is added as a function of wavelength according to external quantum efficiency (EQE ) 53 of TPV (Equation 5-1). EQE is defined as the ratio of the number of electron output by the solar cell to the incident photon of a given wavelength .126 Therefore, the additional reflectance avoids radiation in EnergyPlus simulation which corresponds to the same amount of energy converted by TPV, and the modified optical properties for EnergyPlus simulation is shown in Figure 5-4. REP () = R ( )+ (1 -(ATPV layer ()+EQE ( )) (5-1) Figure 5-4. Optical properties of transparent photovoltaic as input of EnergyPlus simulation : (a) front illumination (b) back illumination Heating sources used in DOE reference building are mainly electricity and natural gas (Table 4 -2), but the geographical charac teristic s of the sources is not considered in each reference building model. In LCA, the impacts from the source of heating are criti cal since there is different environmental impacts between electricity and nat ural gas based on the study areas. Therefore, the simulation considers an additional modification to refl ect the actual heating source based on the CBECS database from EIA. The s elected number of building s for each study area and their major heating source are summarized in Table 5-3, and the portion of major heating source is allocated 54 to the heating energy demand in EnergyPlus simulation by modifying the energy efficiency of a furnace in HVAC system. Tab le 5-3. Heating source of the office building in study area assessed from CBECS database Heating Source Data source Commercial building 114 House 127 Detroit Los Angeles Phoenix Honolulu # of building (n) 95 103 12 312,625 Electricity 7.4% 46.6% 57.1% 3.5% Natural gas 91.6% 46.6% 35.7% 33.5% Others 1.0% 6.8% 7.2% 63.0% 5.3 Result and Discussion 5.3.1 Simulation for Building Energy Demand and Photovoltaic Two different buildings models are used as illustrated and described in Table 4-1. The medium office model is used for the window application while the small office model is used for skylight application. The medium office building has 68% WWR, and the small office has 15 % SRR where TPV are applied as the window and skylight respectively. For HVAC system of b oth buildings , a heating appliance is assumed to use both natural gas and electricity to account for regional differences while air conditioning equipment use s only electricity as shown in Table 5-3. Figure 5-5 summarizes the annual building energy demand and the electricity production for TPV for both window and skylight application. For the climate dependence of energy demand, most of the energy demand in Detroit is from heating while the other three locations require cooling demand dominant ly. The energy demand from fans is not sensitive to climate effect because both heating and cooling need to be distributed all over the building. In term s of building type, there is more energy demand in medium office because of larger size compared to the small office , whereas the trends of heating and cooling demands show a discrepancy . Medium office requires more 55 heating than the small office, and it is presumed to be because of lower heat transfer efficiency from the larger building size and HVAC system . Meanwhile, the portion of fan energy demand is larger in the small office. The loading of fan is increased because the attic space is integrated with the office area to simulate effects of TPV skylight regarding energy demand of HVAC system as mentioned in the Methods section . Figure 5-5. Monthly energy demand for baseline building and electricity producti on by TPV application . Baseline building is installed double pane window, and power conversion efficiency of TPV for electricity production is 10 %. The dotted line with square symbol s for electricity production represents the SEWN application compared to the SEW as shown as a solid line with triangle symbol s in Figure 5-5. Los Angeles, Phoenix, and Honolulu have a similar trend of energy demand with dominant cooling demand, but the ele ctricity production from TPV is similar between Los Angeles and Honolulu 56 and higher in Phoenix because of the latitude. The energy production in Detroit is significantly lower during the winter, but the addition of TPV on the north side increases electrici ty production during the summer. 5.3.2 Assessment of Building Energy Demand by TPV Application The a nnual energy saving for TPV application in window s is shown in Figure 5-6(a) . Both glass and plastic TPV reduce the energy demand from electricity and natural ga s use. Depending on the study area, the portion of energy saving varies due to climate and the source of heating and cooling equipment. In Detroit, annual energy saving is lower than other cities because there is only a smaller opportunity for electricity saving by TPV application because of the cold climate condition that consume s less electricity for air conditioning in summer. However, natural gas saving in winter allows energy saving in Detroit to be comparable to other cities. It implies an internal heat reflection of TPV window could save heating energy demand. Figure 5-6. Annual energy saving of TPV application in window for medium office: (a) glass TPV (b) plastic TPV , with application of TPV in SEWN (south , east, west, and north ) direction 57 For plastic TPV, up to 4% of electricity saving is reduced due to the placement of TPV inside the building as shown in Figure 5 -6 (b). When TPV is placed inside the building , the portion of solar radiation that is absorbed by TPV but not converted to electricity is re -emitted into the building, and this results in more cooling energy demand compared to glass TPV. In Figure 5 -6 (a) and (b), the decrement of electricity saving in warmer climate s (Los Angeles, Phoenix, and Honolulu) is lower than the decrement in cold climate (Detroit ). Overall, energy saving from plastic TPV application is still more than 10% of total energy consumption. Figure 5-7. Annual energy saving of TPV application in skylight for small office (a) glass TPV (b) plastic TPV For the skylight application of TPV as shown in Figure 5 -7, electricity demand is reduced in all locations, but natural gas consumption increases in Detroit. The difference in electricity saving between the glass and plastic TPV applications differs by abo ut 5%. Although net energy saving is positive, the increased use of natural gas lowers the benefit of skylight TPV in Detroit. It shows the trade -off between PV electricity generation and building energy saving. ASHRAE standard 90.1 , The Energy Standard Fo r Buildings Except Low -Rise Residential Building, 58 recommends that the SRR be less than 5 % because of energy conservation efficiency. 128 The ratio is intended to maximize electricity production for this case study, but care need s to be taken under cold climate to make sure both energy saving and electricity generation are consid ered to optimize SRR . 59 Chapter 6 TPV Application in Urban Area: Urban Heat Island Effect and Its Consequential Environmental Impact on Building Sector This work is an extension from the previous chapter to assess the specific advantage of TPV in an urban area. The same building model and TPV are extended to an urban weather simulation. A key point of this chapter is that the energy saving from TPV application contributes to urban climate improvement by reducing heat from the HVAC wasted to the urban energy system. Meanwhile, there is additional energy saving that could be achieved by TPV application under the elevated temperature due to human activities including building in an urban area compared to a rural application of TPV. It suggests that TPV a pplication in an urban area does not create a dverse impacts. Reduction of waste heat from building s would be considered a net environmental benefit although it is not directly quantified in an LCA study. 6.1 Introduction Currently, 54 % of the population lives in urban areas in the world , and it will increase to 66% by 2050 .129 Over 50% of the total energy consumption is from the maintenance and operation of building in high -income cities. 8 BIPV could provide renewable electricity with minimal land use . As discussed in the previous chapter, transparent photovoltaics (TPV) allow to substitute building window and perform similar ly to heat management film also known as low -e film. With TPV application in the medium office, about 10% of the internal energy demand could be saved from HVAC operation. Energy use in an urban area , such as from building operation, is also related to microc limate through the urban energy balance. Total energy gain within the urban area is the sum of waste heat from anthropogenic activities and solar radiation. The excessive waste heat 60 causes urban heat island effect (UHI), which increase the urban temperature by around 10°C to 15°C during daytime .130 UHI can potentially affect communities by increasing energy demand during summertime peak, air conditioning costs, air pollutio n, greenhouse gas emissions, etc. 20 The building performance can also be affected by UHI which would further increase the cooling energy demand by up to 15% under mid -latitude climate in the U.S. because of the warmer temperature .131 Due to this reason, b uilding standards such as ASHRAE Standard 90.1 could be inefficient since thermal insulation requirement s of the standard s underestimate UHI effect .131 TPV application reduce s cooling energy demand by absorbing NIR so that the net energy benefit of building un der urban climate could have even more benefi ts than what was reported in Chapter 4. To assess local climate, a simulation coupling urban climate and anthropogenic activities is essential. Urban weather generator (UWG ) was developed by Massachusetts Institute of Technology in 2011 . The function of the model is to simulate urban climate from rural climate by incorporating urban characteristics. The model can be used alone for assessment of local climate or also be integ rated into building energy simulations .132 Further effort on developing building energy model in UWG was achieve d in 2016 by including the DOE commercial building data. 133 The data includes sixteen different building types wit h three construction eras (pre -1980 construction, post -1980 construction, and new construction). In term of energy use, the model takes into account operation schedules for lighting, plugged -in electrical equipment, occupancy , and gas equipment. 133 This work assesses the UHI effect by TPV application in the urban area by using UWG and simulate s building energy demand from EnergyPlus with the pre -developed building structure of medium office from the previous chapter. 61 6.2 Method The goal of this task is to assess the simultaneo us impacts of UHI effect and energy benefit by TPV application in an urban area. The overall framework of the urban weather simulation and the scenario of building simulation is illustrated in Figure 6-1. UWG models the energy balance between elements repr esenting particular urban characteristic s, heat flux from human activity, solar radiation, and heat exchange from a rural area to an urban area .132,133 The building geometry and TPV application scenario are integrated into the model according to Table 5-1 and Figure 5-1(c). Figure 6-1. Description of Urban weather Generator (UWG) model ,132 and adoption of DOE referen ce building applied TPV in window (model mechanics are adapted from http://urbanmicroclimate.scripts.mit.edu ). Reference weather data from the weather station is simulated to gener ate urban weather data by reflecting urban characteristics including land use, wind speed, a reflection of solar radiation, traffic, and waste heat from the building . 62 Selected city of th is case study is Los Angeles . It represents a realistic case of UHI 22, and California aims to add additional renewables to cover the energy demand from electricity grid by 2035.106 The assessment period s are in August and December to assess the impacts from heating and cooling energy demand. The input parameters for UWG include urban cha racteristics summarized in Table 6-1. Except for the building type, other parameters use the default value representing characteristics of a large city .132 The density of building s in UWG are classified into three categor ies : high -dense, mid -dense, and low -dense. The density determines the number of building per estimated floor area. The high -dense scenario is selected to assess impacts from TPV application to the windows of 48 buildings. From Chapter 5, clear window and glass TPV application are selected to compare impact in urban and rural area s. The thermal parameters (U -factor and SHGC) for the glass TPV and clear window are calculated from EnergyPlus an d imported as input parameters for UWG simulation (Table 6-1). The rural temperature profile is obtained from Typical Meteorological Year 2 (TMY 2), which is a data set derived from the 1961 -1990 National Solar Rad iation Data Base , at Los Angeles International Airport , and input into the UWG simulation. Table 6-1. Urban Weather Generator Parameters as the input s of simulation for generating urban temperature profiles Urban Characteristics Value Unit Microclimate Parameters 133 Urban Boundary Layer Height - Day 1000 m Urban Boundary Layer Height - Night 50 m Urban Characteristics 133 Average Building Height 12 m Building Density 0.54 - Vertical to Horizontal Ratio 0.48 - Road Albedo 0.08 Building Type Medium office with glass TPV application 2.62/0.60 U-factor/SHGC Medium office with clear double pane window 2.74/0.76 U-factor/SHGC Vegetation Parameters 133 Urban Area Veg . Coverage 0.19 - Latent Fraction of Tree 0.5 - Rural Road Vegetation Coverage 0.8 - 63 From the original UWG model, calculation of the solar heat gain to building s is modified to reflect electricity generation by TPV. It is similar to the modification described in Chapter 5 that avoid ed a certain amount of solar radiation based on external quantum efficiency. In the UWG model, TPV efficiency of 10% is assumed to be excluded in the simulation of building energy balance by adjusting SHGC . 6.3 Result and Discussion 6.3.1 Comparison of Urban and Rural Temperature The temperature variation in Los Angeles in August based on TMY2 (rural) and the simulation result (urban) is shown in Figure 6-2. During the summer season, compared to the urban temperature , the rural temperature is lower by about 3 °C at solar noon (~12:3 0 pm) and less than 1°C at daybreak. TPV application does not change the overall temperature, but small changes are observed at the peak of the day . The urban air temperature with the TPV application is slightly lower than the reference glass window , and it suggest s that solar radiation harvested for electricity generation contribute s to the mitigation of UHI. 64 Figure 6-2. Temperature variation in Los Angeles based on rural scenario (green dot line) and urban : baseline ( solid black line) and urban: TPV (red dash line) scenarios during the summer season for a month (Aug. ) During the winter season, the temperature difference between the rural and urban climate varies more than during the summer season. The daily maximum temperature in the rural location is similar to the urban location while minimum temperature is lower by up to 5 °C as shown in Figure 6-3. The trends in temperature di fference are typical of UHI characteristic du ring winter , since built surface absorbs heat during daytime and releases heat in the night . Therefore, the temperature in the urban location is higher than the rural location. 134 The urban climate change for the TPV application is found to be negligible. 65 Figure 6-3. Temperature variation in Los Angeles based on rural scenario (green dot line) to urban: baseline ( solid black line) and urban: TPV (red dash line) scenarios during the winter season for a month (Dec .) 6.3.2 UHI Intensity in Urban Area and Waste Heat from HVAC System Based on the temperature variation, UHI intensity is calculated and shown in Figure 6-4. The calculation of UHI intensity follows the same methods from the previous work and is determined as the averaged air temperature difference between urban and rural area s.135 In Figure 6-4, area above the line indicating 0 shows the existence of UHI effect. In the summer season, UHI occurs during the whole day, and the most intense time is at solar noon when the sun is at the highest position (Figure 6-4). There is an absence of UHI before noon in winter , and most intense UHI occurs before sunrise. In both season s, the impact from TPV application on UHI effect is found to be negligible. Research reports show that ro oftop solar application in Lo s Angeles shows no adverse impacts on air temperature and UHI effect , and there is a potential benefit of cooling air temperature by up to 0.2 °C. 22 The reported result is similar to results for TPV in window application. 66 Figure 6-4. Urban heat island intensity in Los Ang eles during (a) summer season (Aug.) and (b) winter season (Dec.) In the UWG model, the urban temperature is calculated from rural temperature by additional sensible heat flux generated in the urban area as shown in equation 6-1.133 H= , + ++( ,,+ , ,+ & ,+,+ ,) (6-1) Where, Hurb : urban heat flux Qsens , road : sensible heat from road , per urban area Qtraffic : Sensible heat flux from non -building anthropogenic heat, per urban area Qveg : Sensible heat flux from vegetation , per urban area Qsens , wall : Sensible heat flux from wall , per urban area Qsens , window : Sensible heat flux from window , per urban area Qsens , vent&infil : Sensible heat flux from ventilation/infiltration , per urban area 67 Qsens , waste : Sensible heat flux from waste heat ( including . HVAC, gas) , per urban area Qsens , roof : Sensible heat flux from roof , per urban area N: building types simulated With equation 6-1, the potential of UHI mitigation by TPV application could be explained . TPV application reduce s the energy demand of HVAC including cooling and heating. Figure 6-6 illustrate s the monthly average waste heat from H VAC per urban area. In summer, there is more waste heat because of more cooling energy demand in the daytime . Comparing clear window (baseline ) to TPV application, the reduction of waste heat is higher in the daytime while there is no difference in waste heat after sunset. The s ame trend is shown in winter, but the temporal patterns and magnitude of waste heat are different due to climate. There fore, the concept of waste heat helps to explain the UHI intensity graph (Figure 6-4) showing a little reduction of urban temperature in summer. In this case study, the medium office is only simulated for TPV application in Los Angeles , and it shows that the reduction of waste heat is not enough to lower the urban temperature . However, a comprehensive assessment of UHI mitigation can be achieved if the realistic data regarding type and number of building s are allowed . 68 Figure 6-5. Emission of waste heat from HVAC to the urban area in (a) summer season (Aug.) (b) winter season (Dec.) 6.3.3 Energy Saving by TPV Application in Rural and Urban Area s TPV application in the urban area compared to rural location has more benefits due to warmer temperature caused by UHI. Figure 6-8 shows the monthly energy saving in GJ by TPV application in the rural and urban area. The building energy demand of a single medium office building with TPV application is simulated by using the climate data from both the Los Angeles International Airport (rural characteristics) and the UWG simulation (urban characteristics). There is electricity saving in both summer and winter of up to 2 GJ , from air con ditioning and fan operation s. Since the climate in the urban area is warmer than rural climate during summer as shown in Figure 6-2, there is more saving during the summer. The natural gas saving is negligible because of climate and schedule of heating ope ration based on the regular working hour s for a commercial building. 69 Figure 6-6. Comparison of monthly energy saving for TPV application in window for urban and rural scenarios in Los Angeles during summer (Aug.) and winter (Dec.) seasons In this chapter, the TPV application in the urban area is discussed with microclimate and building energy demand simulation s. TPV application shows no adverse impa cts on air temperature, and there is a potential benefit of cooling the urban air temperature because of the reduced waste heat. It is contrary to a previous study showing the UHI effect induced by a large solar power plant in Arizona. 136 The absorbed sol ar radiation that is converted to electricity reduce s the building energy demand so that the overall energy balance in an urban area results in cooling the air temperature. The building consumes more energy under UHI effect . However , the consumption can be reduced with TPV application by up to 2 GJ/month in the urban area compared to the application in the rural location. 70 Chapter 7 Conclusion s OPV material process is improved by using the iterative evaluation method developed in this work which is based on green chemistry and life cycle approach. The material produced from the new process reduce s the overall impact of OPV device, especially TPV made of ClAlPc. The application for window reduc es the energy demand of building and waste heat from HVAC system, and shows potential for urban climate improvement. These benefits would be considered as net environmental benefit of TPV application and could be quantified with a projection of overall impacts th rough life cycle assessment. In Chapter 2, synthesis pathways of ClAlPc are established by applying green chemistry principles and metrics. The work identifies a greener pathway for ClAlPc synthesis by starting from phthalonitrile as the precursor . The at om economy of phthalonitrile pathway is about 30% higher th an that of phthalic anhydride pathway . The phthalic anhydride pathway has a theoretical yield of 60%, however, because of byproduct generation during intermediate step, the actual yield lowers to 2 5%. The application of microwave reactor to the phthalonitrile pathway reduces reaction time from six hours to 40 minutes that suggests better energy efficiency and yield . Chapter 3 evaluates the performance of OPV devices from ClAlPc synthesized from different processes. Two precursors of ClAlPc, phthalonitrile and phthalic anhydride, are studied to evaluate the tradeoff between material purity, device efficiency, and materi al cost . The results show that UV -vis and mass spectra of ClAlPc from phthalonitrile and phthalic anhydride processes are in agreement with the reference material. The best device from phthalonitrile sample shows similar energy conversion efficiency with t he reference device although the fabrication yield is low. For phthalic anhydride samples, all devices work, but have low energy conversion efficiency 71 due to impurities from urea decomposition. The result s of Chapter 2 and 3 are used as baseline of evaluat ion method developed in Chapter 4. A new methodology is developed by combining LCA and green chemistry metrics to evaluate OPV material process es. Through the experimental work basis of Chapter 2, the hotspot of baseline process of ClAlPc was identified as toxic solvent use and high energy demand. Applying the iterative evaluation method could solve issues related to each intermediate process regar ding environmental, cost, and chemical hazards. Finally, a new process using diethylene glycol under microwave reactor is developed to produce ClAlPc which reduces the environmental impact s by 3%, the cost by 9% and chemicals hazard by 23% simultaneously compared to the baseline process , while producing high -quality material that can be used for OPV application . The application of TPV made of ClAlPc is assessed to evaluate the energy benefit of power generating window. The energy benefit is a key aspect of TPV application, and it will be used for use phase input data in LCA for net environmental benefit of TPV application for window s over its lifetime. The building energy demand simulation of the medium office and the small office identifies energy benefit in locations representing different climate s and environmental impacts from the electricity grid. The reduction of cooling energy demand in all study location and type of applications results in annual energy saving w hile the natural gas demand in Detroit show s contrary trends by application type. The window application saves natural gas while skylight application consumes more natural gas due to heat loss from the skylight . However, overall net energy benefit is notab le for both applications in term of energy saving and PV electricity generation. The energy saving from HVAC by TPV application in window s is seen under cold climate condition as well, even though the cold climate condition, such as Detroit, requires highe r heating energy demand. 72 The last part of the work is related to Chapter 5, but the application consider s not only location but also urban impact as well . TPV application for a window in an urban area could result in additional benefit s in term s of saving land use and reducing waste heat contributing to UHI effect . Also, the energy saving of the TPV application in the urban area is higher than the application in the rural area up to 2 GJ per month. The o verall result implies that TPV window in the urban area reduces energy demand and ha s additional potential for deployment of renewable energy devices and mitigation of UHI effect, leading toward s urban sustainability. TPV application in window studied in chapter 5 and chapter 6 provides motivating research questions for future study. Regarding cooling energy demand, the absorption profile of TPV plays an essential role since NIR absorption reduce s energy demand. Regarding the TPV absorption, the next step in this subject is to study different types of TPV that harvest electricity from different region s of the solar spectrum. The requirement of thermal transmittance of a building is recommended by climate conditions , to save energy in both cooling and heating, and thus there could be optimal configurations of TPV specific to climate conditions . The applications of TPV in window are only limited to office building s in chapter 4 and chapter 5. In chapter 5, waste heat from HVAC could be reduced by TPV application since TPV application aids to reduce energy demand from HVAC used in the office buildings. HVAC systems vary depending on the purpose and type of building. Therefore, the benefit of TPV application from the reduction of waste heat from HVAC will differ . Various types of building exist in an urban area , and application of TPV to each building will result in benefit of differing magnitude and temporal patterns. Another study for TPV application will be to study TPV in window s of all the different DOE commercial reference buildings to assess the patterns of energy saving and reduction of waste heat from HVAC. 73 APPENDIX 74 APPENDIX 1. Summary of Evaluation Methods for Chemicals In Table A 1, 19 studies regarding evaluation method for chemical process are summarized by their principles, method, and application area, and evaluation criteria. Most studies cover environmental impact while either LCA or own metrics partially cover health and cost impact s. The purpose of application widely covers from lab scale experiment to industrial application in addition to educational purpose and synthesis plan. The methods for lab scale experiment are based on point system or own metrics 137Œ139 for simplicity. Screening methods aim to develop chemical process in early stage so that wide evaluatio n criteria are considered depending on own purpose of application while the system boundary of evaluation is generally limited in the process (gate to gate) similar to the lab scale evaluation. Among the studies considering green chemistry principles, only few studies include cost analysis 137,140 Œ142, and it is presumed due to the reason that the 12 Green Chemistry principles do not explicitly include the economic aspect. The evaluation approach aims to fine chemicals process that involve complex organic molecules and requires hig h purity. Therefore, it is important to include our system boundary from cradle to gate so that the evaluation method could identify and prevent unintended impacts associated with environmental, cost and chemical hazard from upstream process with iterative approach. 75 Table A 1. Summary of existing method for fine chemical evaluation Paper Principle Method Purpose of Application Evaluation criteria Ref. Our method LCA; Green chemistry Iterative approach based on streamlined LCA and green chemistry metrics Screening method for fine chemicals CED; Global warming potential; water demand; chemical and electrical cost; E -factor NFPA 704 - Koller et al., 2000 EHS (Environmental, Health, and Safety) Mass balance Screening method for fine chemicals Safety; Health; Environment 143 Saling et al., 2002 (Eco -efficiency, BASF) LCA LCA (Cradle to grave) Cost analysis Industrial application for manufacturing CED; Emission; Material depletion; Risk; Toxicity; cost 64 Eissen and Metzger, 2002 (EATOS ) Green chemistry Mass balance Screening method Mass index; E -factor 144 Gonzalez and Smith, 2003 (GREENSCOPE, EPA) Green chemistry Own metrics based on process Bench to process level application for decision making Efficiency (material); Environment; Energy; Economics 140 Aken et al., 2006 (EcoScale) Green Chemistry Penalty points given in evaluation criteria Lab scale evaluation Yield; Cost; Safety; Technical steps (reaction, work -up, purification) 137 Curzons et al., 2007 (FLASC, GSK) LCA; EHS LCA (Cradle to gate) Hierarchical Cluster Analysis (HCA) Principle Component Analysis (PCA) Industrial application for manufacturing Energy; GHGs; Fossil fuel; Acidification Eutrophication; photochemical ozone creation potential; Total organic carbon 65 Ribeiro et al., 2010 (Green Star) Green chemistry Score given in evaluation criteria Lab scale evaluation Score for each 12 Green chemistry principles 138 Mercer et al., 2012 (Multivariate Method) Green chemistry; LCA Own metrics based on chemical reaction Lab scale evaluation for education purpose Acidification; Ozone depletion; Smog formation; Global warming; Human toxicity; Persistence; Bioaccumulation; Abiotic Depletion 139 Andraos, 2012 (Benign index, CareerChem ) Green Chemistry Own metric (BI) based on process Screening method for chemical reaction and synthesis plan Acidification -basification; Ozone depletion; Smog formation; Global warming; Inhalation toxicity; Ingestion toxicity; Bioconcentration; Abiotic resource 145 Russell and Shiang, 2012 (DCSFT) LCA; Questions Score based on justification Planning for industrial application Economic; Social; Resource use; Water; Green House Gases; Organization 66 Andraos, 2013 (Safety/Hazard Indices, CareerChem) Safety; Hazard Own metric (SHI) based on chemical properties Screening method for chemical reaction and synthesis plan Flammability; corrosiveness (gas and liquid); oxygen balance; hydrogen generation; explosive vapor; explosive strength; impact sensitivity; risk phrase; occupational exposure limit; dermal absorption; skin dose 146 McElroy et al., 2015 (CHEM21) Green chemistry; LCA; EHS Score based on GHS and chemical properties Methodology for bench scale to manufacturing Evaluation criteria considering sustainability varies up to Each stage 142 76 Table A1 (cont™d) Eckelman, 2016 Green Chemistry LCA LCA (Cradle to gate ) Own metric (inherent toxicity, i*) Screening method for inherent chemical toxicity Ingestion non -cancer and cancer; inhalation non -cancer and cancer 147 Alder et al., 2016 (GSK™s solvent sustainability guide) Green Chemistry; LCA; EHS LCA (Cradle to gate ) Score based on GHS and chemical properties Solvent selection guide Boiling point; incineration; recycling; biotreatment; VOC emission; aquatic impact; air impact; health hazard; exposure potential; safety; LCA 84 Mercado et al., 2016 (GREENSCOPE extension) Green Chemistry; EHS Synergistic approach of WAR Algorithm, GREENSCOPE. and SustainPro tools Industrial application for sustainability assessment GREENSCOPE : Efficiency (material); Environment; Energy; Economics WAR Algorithm : ozone depletion, global warming, smog formation, acid rain formation, human toxicity, ecotoxicity SustainPro : energy consumption, waste of material, costs, capacities, time, and efficiency 141 Smith et al., 2017 Green Chemistry; LCA LCI (gate to gate) Life cycle inventory analysis of chemical manufacturing Emission; energy; mass 148 Kreuder et al., 2017 (CGM, MilliporeSigma) Green chemistry Green Chemistry metrics; own metrics Screening method for industrial chemicals Resource use; Energy efficiency; Human and environmental hazard 149 Song et al., 2017 LCA QSAR; Machine learning Screening method CED; Acidification; GHGs; Eco -indicator 99; Human health; Ecosystem 150 iSUSTAINŽ Green Chemistry Index v2.0 Green Chemistry Green Chemistry metrics; own metrics; Industrial application for sustainability assessment Green chemistry index within 12 Green Chemistry principles 151 2. Detailed Methodology 2.1. ClAlPc Synthesis Baseline and five alternative processes to demonstrate the iterative process based on evaluation of sustainability are introduced in the main chapter (Table 1 -1). However, detailed information regarding to upstream process are omitted so that Table A 2-A5 illustrates the material flow and resource use within system boundary. White boxes are new inventory data, and gray boxes correspond to information from existing inventory data while black boxes are direct emission from process. 77 The most common precursor of phthalocyanine among phthalic acid and its derivates is 1,2 -Dicyanobenzene (phthalonitrile, PN), and PN is produced from o-xylene under ammoxidation process at -85%. 34 Another precursor determining a type of metal phthalocyanine is metal salts, and thus aluminum chloride anhydrous (AlCl 3) is used to synthesis chloroaluminum phthalocyanine (ClAlPc) in our experiment. AlCl 3 are produced by chlorinating molten aluminum at 670 - 35Other information of chemical production is summarized in Table A 2. The reaction of ClAlPc from phthalonitrile (PN) and phthalic anhydride (PA) is carried out using either heating mantle or a microwave reactor (CEM Corporation Discover SP). In ClAlPc synthesis, the molar ratio of PN or PA to aluminum ion is fixed to 4:1. 33,36,37 AlCl 3 and PN or PA are mixed with the metal salt in reaction media. High boiling point solvent is required for all 78 Figure A 1. Flow diagram of the baseline process Figure A 2. Flow diagram of P1 79 Figure A 3. Flow diagram of P2 Figure A 4. Flow diagram of P3 80 Figure A 5. Flow diagram of P4 81 Figure A 6. Flow diagram of P5 82 2.2. Material Characterization UV-vis spectrometry and HPLC -MS are used to validate purity of ClAlPc samples for OPV application. Samples are dissolved in ethanol with 15 min of sonication and filtered through 22 µm syringe filter 60, and reference (TCI America 98 %) is compared to the samples. All samples have the same maxima in the UV -vis spectrum with the reference at 670 nm although spectrum maxima of P1 is at 675 nm that is due to acidity of sample. 60 HPLC -MS results show more informative chemical identification since impurities of baseline, P2, and P3 are identified at different m/z and retention time. As shown in the Tabl e A9, HPLC -MS spectrum at 7.13 min has peak at 614.12 m/z, and it implies that additional chlorine ion is in chemical structure by attack by a nucleophile. 38,55 Meanwhile, P2 and P3 have peak at 487.17 m/z, and the common feature of both processes is the use of urea as nitrogen source. Urea decomposition is previously studied and known as that the thermal decomposition at around 200 61 Therefore, it is fair assumption that the derivatives remain after the separation process of ClAlPc. 83 Figure A 7. UV-vis spectra of ClAlPc samples 84 Figure A 8. HPLC -MS spectra of ClAlPc samples (at 6.81 min) 85 3. Data Source for Evaluation Methods 3.1. Materials for Life Cycle Assessment Table A 2. Data sources for materials and energy used in the inventory analysis Details Ref. 1,2-dicyanobenzene Ammoxidation process. o -xylene is converted to 1,2 -dicyanobenzene by reaction with gaseous ammonia and oxygen 34 Aluminum chloride Chlorination of molten aluminum with chlorine gas 35 Naphthalene Refining process of coal tar 152 1-chloronaphthalene Stoichiometric reaction equation 77,78 Anisole Methylation of sodium p henoxide with dimethyl sulfate 153 Sodium phenoxide Phenol is converted to sodium phenoxide by reaction with sodium hydroxide 153 2,4-dichloroanisole Chlorination of anisole by reaction with sodium chlorate 154 Caprolactam Oxidation of cyclohexan e process in China 155 1,8-Diazabicyclo (5.4.0)undec -7-ene Synthesis of bicyclic amidines from caprolactam by using acrylonitrile 156 Utility input for reaction and workou t Utility input for chemicals with no specific information is assumed to be 1.2 kg/kg of steam, 0.7 MJ/kg of electricity, 70 kg/kg of cooling water, and 0.06 Nm 3/kg of nitrogen gas per kg of product 78 Utility input for solvent regeneration Utility input for solvent regeneration is assumed to be 1.5 kg/kg of steam, 0.2 MJ/kg of electricity, 80 kg/kg of cooling water, and 0.01 Nm 3/kg of nitrogen gas per kg of produ ct 78 Transportation (chemical) Average distance for basic chemicals in the U.S: 69 km truck, 43 km train, barge 27 km, and pipeline 18km 157 Transportation (nonflammable gas) Average distance for gases in the U.S: 107 km truck, 8 km train, barge 6 km, and pipeline 5 km 157 86 3.2. Chemical Identification and Hazards from NFPA Table A 3. Chemical identification and NFPA score of materials used in ClAlPc processes No. Chemicals CAS Health Flammability Reactivity Special 1 1,8 -Diazabicyclo(5.4.0)undec -7-ene 6674 -22-2 3 1 0 0 2 1-chloronaphthalene 90-13-1 2 1 0 0 3 2,4 -Dichloroanisole 553 -82-2 0 1 0 0 4 Acetic acid 64-19-7 3 2 0 0 5 Acetone 67-64-1 2 3 0 0 6 Acrylonitrile 107 -13-1 3 3 0 0 7 Alumina 1344 -28-1 2 0 0 0 8 Aluminum 7429 -90-5 0 3 3 0 9 Aluminum chloride 7446 -70-0 3 0 2 0 10 Ammonia, liquid 7664 -41-7 3 0 0 0 11 Ammonium chloride 12125 -02-9 2 0 0 0 12 Anisole 100 -66-3 0 2 0 0 13 Bauxite 1318 -16-7 1 0 0 0 14 Benzene 71-43-2 2 3 0 0 15 Bituminous coal 2684C (SRM) 1 3 0 0 16 Caprolactam 105 -60-2 2 1 0 0 17 Chlorine, liquid 7782 -50-5 3 0 0 1 18 Chloroaluminum phthalocyanine 14154 -42-8 0 0 0 0 19 Coal tar 1991 (SRM) 1 1 0 0 20 Crude oil 2721 (SRM) 2 3 0 0 21 Diesel 68334 -30-5 1 2 0 0 22 Diethylene glycol 111 -46-6 0 1 0 0 23 Dimethyl sulfate 77-78-1 4 2 0 0 24 Ethyl acetate 141 -78-6 2 3 0 0 25 Fuel grade uranium (Uranium dioxide) 1344 -57-6 3 0 0 0 26 Gasoline 8032 -32-4 1 3 0 0 27 Hydrochloric acid 7647 -01-0 3 0 0 0 28 Hydrogen peroxide 7722 -84-1 3 0 2 1 29 Hydrogen, liquid 1333 -74-0 3 4 0 0 30 Hydroquinone 123 -31-9 2 1 0 0 31 Lignite coal 1415 -93-6 1 1 1 0 32 Limestone 1d (SRM) 1 0 0 0 33 Liquefied petroleum gas (propane) 74-98-6 2 4 0 0 34 Metallurgical coke 65996 -77-2 1 1 0 0 35 Methanol 67-56-1 2 3 0 0 36 Naphthalene 91-20-3 2 2 2 0 87 Table A 3 (cont™d) 37 Natural gas 68410 -63-9 1 4 0 0 38 Nickel 7440 -02-0 2 3 3 0 39 Nitric acid 7697 -37-2 3 0 2 1 40 Nitrogen, liquid 7727 -37-9 3 0 0 1 41 Oxygen, liquid 7782 -44-7 3 0 0 1 42 Petroleum coke 64741 -79-3 1 1 0 0 43 Petroleum refining (Petroleum ether) 8032 -32-4 2 4 0 0 44 Phenol 108 -95-2 3 2 0 0 45 Phthalic anhydride 85-44-9 2 1 0 0 46 Phthalonitrile 91-15-6 2 1 0 0 47 Potassium chloride 7447 -40-7 0 0 0 0 48 Pyrolysis gasoline 68606 -10-0 1 3 0 0 49 Quicklime (Calcium oxide) 1305 -78-8 3 0 0 0 50 Residual fuel oil 68476 -33-5 2 2 0 0 51 Sodium chlorate 7775 -09-9 1 0 2 1 52 Sodium chloride 7647 -14-5 1 0 0 0 53 Sodium hydroxide 1310 -73-2 2 0 0 0 54 Sulfur 7704 -34-9 2 1 2 0 55 Sulfuric acid 7664 -93-9 3 0 0 0 56 Toluene 108 -88-3 2 3 0 0 57 Urea 57-13-6 0 0 0 0 58 Xylene 1330 -20-7 2 3 0 0 Table A 4. Chemical hazard weighed to material consumption for ClAlPc manufacturing based on input from technosphere Process Health Flammability Reactivity Special Baseline 1.39 3.12 0.06 0.03 P1 1.66 2.70 0.11 0.05 P2 1.62 2.97 0.05 0.02 P3 1.77 2.95 0.02 0.01 P4 1.76 2.97 0.02 0.01 P5 1.53 2.84 0.03 0.03 88 Table A 5. Chemical hazard weighed to material consumption for ClAlPc manufacturing based on input from tech nosphere except energy source (grey in Table A2.) Process Health Flammability Reactivity Special Baseline 2.07 2.28 0.20 0.09 P1 2.07 2.29 0.20 0.09 P2 2.01 2.61 0.08 0.03 P3 2.18 2.47 0.03 0.02 P4 2.18 2.47 0.03 0.02 P5 1.89 2.24 0.06 0.05 3.3. Summary of Green Chemistry Metrics and Solvents Use Table A 6. Green chemistry metrics of proposed processes and solvents use Process AE % RME % PMI E-factor E-factor (water use) Yield % Solvent use (g) Baseline 89.02 74.89 46.45 8.45 16.69 84.20 Toluene Acetone Water 31.22 29.81 14.00 P1 89.02 64.76 53.88 9.93 19.45 72.80 Toluene Acetone Water 31.22 29.31 14.00 P2 59.52 9.85 308.56 43.68 729.19 24.80 Acetone Water H2SO4 NaOH 79.20 195.37 1.86 0.40 P3 89.02 36.52 487.45 38.77 357.66 50.54 Acetone Water H2SO4 NaOH 79.2 186.20 9.80 4.00 P4 89.02 43.83 406.09 32.12 297.71 60.66 Acetone Water H2SO4 NaOH 79.2 186.20 9.80 4.00 P5 89.02 62.63 29.16 8.11 8.11 70.72 Methanol 31.65 3.4. Cost of Crude Oil, Coals, and Natural Gas The price of petroleum, coal, and natural gas from life cycle inventory to consider the cost of chemicals and energy. The reference prices of fossil fuels are retrieved from Energy Information Administration (EIA) for the average of most recent data ( Table A 7). For C 2, the production rate of ClAlPc is modeled on either a pilot scale conventional reactor (20L) or microwave rea ctor (3.5L) based on previous studies ( Table A 8). 158,159 Comparing the production rate is important because it provides an estimate of use and maintenances of facility. 89 Table A 7. Reference cost of crude oil, coals, and natural gas unit Price ($) Period Crude oil kg 0.41 09/17 -02/18 Bituminous kg 0.05 01/16 -12/16 Lignite kg 0.02 01/16 -12/16 Anthracite kg 0.10 01/16 -12/16 Gas natural m3 0.14 09/17 -02/18 Table A 8. Production rate of baseline and alternative processes Reactor Production rate (kg/hr) Baseline Pilot 20L 0.41 P1 Microwave 3.5L 0.56 P2 Microwave 3.5L 0.24 P3 Microwave 3.5L 0.60 P4 Microwave 3.5L 0.73 P5 Microwave 3.5L 0.58 3.5 Sustainability Score Summary Table A 9. Summary of sustainability scores for all methods Process Environmental (%) Cost (%) Chemical hazard (%) Baseline 100 100 100 P1 131 100 108 P2 613 84 408 P3 893 77 352 P4 733 77 284 P5 97 77 91 90 REFERENCES 91 REFERENCES (1) Keoleian, G. A.; Lewis, G. M. Modeling the life cycle energy and environmental performance of amorphous silicon BIPV roofing in the US. Renew. Energy 2003, 28 (2), 271Œ293. (2) Cho, A. Energy™s Tricky Tradeoffs. Science (80 -. ). 2010, 329 (5993), 786 Œ787. (3) Anctil, A.; Babbitt, C. W.; Raffaelle, R. P.; Landi, B. J. Cumulative energy demand for small molecule and polymer photovoltaics. Prog. Photovoltaics Res. Appl. 2013, 21 (7), 1541Œ1554. (4) Heliatek sets new world record efficiency of 10.7 % for its organic tandem cell https://www.heliatek.com/en/press/press -releases/details/heliatek -sets -new -world -record -efficiency -of-107-for -its-organic -tandem -cell (accessed Mar 7, 2018). (5) U.S. Environmental Protection Agency. Mercury and air t oxics standards https://www.epa.gov/mats/cleaner -power -plants (accessed Jul 11, 2018). (6) EIA. Annual energy outlook 2018 ; 2018. (7) Attoye, D. E.; Aoul, K. A. T.; Hassan, A. A review on building integrated photovoltaic façade customization potentials. Sustain. 2017, 9 (12). (8) UN-Habitat. State of the world™s cities 2008 -2009: harmonious cities ; Earthscan: London, 2008. (9) Internat ional Energy Agency. Transition to Sustainable Buildings: Strategies and Opportunities to 2050 ; Paris, 2013. (10) International Energy Agency. Energy Technology Perspectives 2016 ; Paris, 2016. (11) Traverse, C. J.; Pandey, R.; Barr, M. C.; Lunt, R. R. Em ergence of highly transparent photovoltaics for distributed applications. Nat. Energy 2017, 2 (11), 849 Œ860. (12) Anctil, A.; Babbitt, C. W.; Raffaelle, R. P.; Landi, B. J. Material and energy intensity of fullerene production. Environ. Sci. Technol. 2011 , 45, 2353Œ2359. (13) Anastas, P.; Eghbali, N. Green Chemistry: Principles and Practice. Chem. Soc. Rev. 2010 , 39 (1), 301 Œ312. (14) Calvo -Flores, F. G. Sustainable chemistry metrics. ChemSusChem 2009, 2, 905Œ919. (15) Burke, D. J.; Lipomi, D. J. Green chemistry for organic solar cells. Energy Environ. Sci. 2013, 6, 2053. (16) Pearce, J. M. Optimizing Greenhouse Gas Mitigation Strategies To Supress Energy 92 Cannibalism. In 2nd Climate Change Technology Conference ; Hamilton, Ontario, Canada, 2009; pp 1Œ9. (17) Bhandari, K. P.; Collier, J. M.; Ellingson, R. J.; Apul, D. S. Energy payback time (EPBT) and energy return on energy invested (EROI) of solar photovoltaic systems: A systematic review and meta -analysis. Renew. Sustain. Energy Rev. 2015, 47, 133Œ141. (18) Chae, Y. T.; Kim, J.; Park, H.; Shin, B. Building energy performance evaluation of building integrated photovoltaic (BIPV) window with semi -transparent solar cells. Appl. Energy 2014, 129, 217Œ227. (19) International Energy Agency. Cities are at the frontline of the energy transition https://www.iea.org/newsroom/news/2016/september/cities -are -at-the -frontline -of-the -energy -transition.html. (20) Heat Island Effect https://www.epa.gov/heat -islands. (21) Wang, Y.; Tian, W.; Zhu, L.; Ren, J.; Liu , Y.; Zhang, J.; Yuan, B. Interactions Between Building Integrated Photovoltaics and Microclimate in Urban Environments. Sol. Energy 2005, 2005 (May 2006), 499 Œ504. (22) Taha, H. The potential for air -temperature impact from large -scale deployment of sola r photovoltaic arrays in urban areas. Sol. Energy 2013 , 91, 358Œ367. (23) Sheldon, R. a. Fundamentals of green chemistry: efficiency in reaction design. Chem. Soc. Rev. 2012, 41 (4), 1437. (24) Cespi, D.; Beach, E. S.; Swarr, T. E.; Passarini, F.; Vassur a, I.; Dunn, P. J.; Anastas, P. T. Life cycle inventory improvement in the pharmaceutical sector: assessment of the sustainability combining PMI and LCA tools. Green Chem. 2015, 17 (6), 3390 Œ3400. (25) Curzons, A. D.; Constable, D. C.; Cunningham, V. L. S olvent selection guide: a guide to the integration of environmental, health and safety criteria into the selection of solvents. Clean Prod. Process. 1999 , 1 (2), 82 Œ90. (26) Watson, W. J. W. How do the fine chemical, pharmaceutical, and related industries approach green chemistry and sustainability? Green Chem. 2012, 14 (2), 251. (27) Mulvihill, M. J.; Beach, E. S.; Zimmerman, J. B.; Anastas, P. T. Green Chemistry and Green Engineering: A Framework for Sustainable Technology Development. Annu. Rev. Environ. Resour. 2011, 36 (1), 271 Œ293. (28) Anctil, A.; Fthenakis, V. Life Cycle Assess ment of Organic Photovoltaics. In Third Generation Photovoltaics ; InTech, 2012. (29) Curran, M. A. Life cycle assessment: principle and practice ; Cincinnati, OH, 2006. (30) Lunt, R. R.; Bulovic, V. Transparent, near -infrared organic photovoltaic solar ce lls for 93 window and energy -scavenging applications. Appl. Phys. Lett. 2011, 98 (11), 3 Œ5. (31) Young, M.; Suddard -Bangsund, J.; Patrick, T. J.; Pajares, N.; Traverse, C. J.; Barr, M. C.; Lunt, S. Y.; Lunt, R. R. Organic Heptamethine Salts for Photovoltaics and Detectors with Near -Infrared Photoresponse up to 1600 nm. Adv. Opt. Mater. 2016, 4 (7), 1028 Œ1033. (32) Salzman, R. F.; Xue, J.; Rand, B. P.; Alexander, A.; Thompson, M. E.; Forrest, S. R. The effects of copper phthalocyanine purity on organic solar cell performance. Org. Electron. physics, Mater. Appl. 2005 , 6 (5Œ6), 242 Œ246. (33) Löbbert, G. Phthalocyanin es. In Ullmann™s Encyclopedia of Industrial Chemistry ; Wiley -VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2000. (34) Lorz, P. M.; Towae, F. K.; Enke, W.; Jäckh, R.; Bhargava, N.; Hillesheim, W. Phthalic Acid and Derivatives. In Ullmann™s Encyclopedia of Industrial Chemistry ; Wiley -VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2007; pp 35 Œ154. (35) Helmboldt, O.; Keith Hudson, L.; Misra, C.; Wefers, K.; Heck, W.; Stark, H.; Danner, M.; Rösch, N. Aluminum Compounds, Inorganic. In Ullmann™s Encyclopedia o f Industrial Chemistry ; Wiley -VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2007. (36) McKeown, N. B. The Synthesis of Symmetrical Phthalocyanines. In The Porphyrin Handbook ; Elsevier, 2003; Vol. 15, pp 61 Œ124. (37) Erk, P.; Hengelsberg, H. Phthalocyani ne Dyes and Pigments. In The Porphyrin Handbook ; Elsevier, 2003; pp 105 Œ149. (38) Yasuhiro, Y.; Kenji, T. Mu -oxo Bridged Heterometal Compound And Selective Production Method Thereof. EP 1 418 207 A1, 2004. (39) Guay, D.; Dodelet, J. P.; Cote, R.; Langfor d, C. H.; Gravel, D. Photoelectrochemistry of some aluminum phthalocyanines in regenerative solar cells. J. Electrochem. Soc. 1989, 136 (8), 2272 Œ2280. (40) Owen, J. E.; Kenney, M. E. Phthalocyaninoaluminum Compounds. Inorg. Chem. 1962, 1 (2), 331 Œ333. (41) Sheldon, R. a. E factors, green chemistry and catalysis: an odyssey. Chem. Commun. (Camb). 2008 , No. 29, 3352 Œ3365. (42) Vincoli, J. W. Risk management for hazardous chemicals ; CRC Press, 1997. (43) Burczyk, A.; Loupy, A.; Bogdal, D.; Petit, A. Imp rovement in the synthesis of metallophthalocyanines using microwave irradiation. Tetrahedron 2005, 61 (1), 179 Œ188. (44) Shaabani, A.; Maleki -Moghaddam, R.; Maleki, A.; Rezayan, A. H. Microwave assisted synthesis of metal -free phthalocyanine and metalloph thalocyanines. Dye. Pigment. 2007 , 74 (2), 279 Œ282. 94 (45) Roy, M. S.; Balraju, P.; Deol, Y. S.; Sharma, S. K.; Sharma, G. D. Charge -transport and photocurrent generation in bulk hetero junction based on Chloro -aluminum phthalocyanine (ClAlPc) and Rose Beng al (RB). J. Mater. Sci. 2008, 43 (16), 5551 Œ5563. (46) -Oxo bridged heterometal compound and selective production method thereof. 03025734.9, 2004. (47) de la Hoz, A.; Díaz -Ortiz, Á.; Moreno, A. Microwaves in organic synthesis. Thermal and non-thermal microwave effects. Chem. Soc. Rev. 2005, 34 (2), 164 Œ178. (48) Rodenburg, L.; Guo, J.; Christie, R. Polychlorinated biphenyls in pigments: Inadvertent production and environmental significance. Color. Technol. 2015 , 131 (5), 353 Œ369. (49) Hu, D.; Hornbuckle, K. C. Inadvertent polychlorinated biphenyls in commercial paint pigments. Environ. Sci. Technol. 2010 , 44 (8), 2822 Œ2827. (50) Lunt, R. R. Theoretical limits for visibly transparent photovoltaics. Appl. Phys. Lett. 2012 , 101 (4), 043902. (51) Bailey -Salzman, R. F.; Rand, B. P.; Forrest, S. R. Near -infrared sensitive small molecule organic photovoltaic cells based on chloroaluminum phthalocyanine. Appl. Phys. Lett. 2007, 91 (2007), 1 Œ4. (52) Harada, K.; Edura, T.; Adachi, C. N anocrystal Growth and Improved Performance of Small Molecule Bulk Heterojunction Solar Cells Composed of a Blend of Chloroaluminum Phthalocyanine and C 70. Appl. Phys. Express 2010, 3 (12), 121602. (53) Zheng, Y. -Q.; Zhang, J.; Yang, F.; Komino, T.; Wei, B.; Zhang, J.; Wang, Z.; Pu, W.; Yang, C.; Adachi, C. Influence of deposition substrate temperature on the morphology and molecular orientation of chloroaluminum phthalocyanine films as well the performan ce of organic photovoltaic cells. Nanotechnology 2015, 26 (40), 405202. (54) Li, N.; Forrest, S. R. Tilted bulk heterojunction organic photovoltaic cells grown by oblique angle deposition. Appl. Phys. Lett. 2009, 95 (12), 123309. (55) Christie, R. M.; De ans, D. D. An investigation into the mechanism of the phthalonitrile route to copper phthalocyanines using differential scanning calorimetry. J. Chem. Soc. Perkin Trans. 2 1989, No. 2, 193. (56) Guay, D. Photoelectrochemistry of Some Aluminum Phthalocyani nes in Regenerative Solar Cells. J. Electrochem. Soc. 1989, 136 (8), 2272. (57) Sigma Aldrich http://www.sigmaaldrich.com/united -states.html. (58) TCI America http://www.tcichemicals.com/en/us/index.html. (59) Molbase http://www.molbase.com/en/index.php. 95 (60) Ghani, F.; Kristen, J.; Riegler, H. Solubility Properties of Unsubstituted Metal Phthalocyanines in Different Types of Solvents. J. Chem. Eng. Data 2012, 57 (2), 439 Œ449. (61) Schaber, P. M.; Colson, J.; Higgins, S.; Thielen, D.; Anspach, B.; Brauer, J. Thermal decomposition (pyrolysis) of urea in an open reaction vessel. Thermochim. Acta 2004, 424 (1Œ2), 131 Œ142. (62) Pollak, P.; Vouillamoz, R. Fine Chemicals. In Ullmann™s Encyclopedia of Industrial Chemistry ; Wiley -VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2013; pp 37 Œ49. (63) Wernet, G.; Conradt, S.; Isenring, H. P.; Jiménez -González, C.; Hungerbühler, K. Life cycle assessment of fine chemical production: A case study of pharmaceutical synthesis. Int. J. Life Cycle Assess. 2010, 15 (3), 294 Œ303. (64) Saling, P.; Kicherer, A.; Dittrich -Krämer, B.; Wittlinger, R.; Zombik, W.; Schmidt, I.; Schrott, W.; Schmidt, S. Eco -efficiency analysis by basf: the method. Int. J. Life Cycle Assess. 2002, 7 (4), 203 Œ218. (65) Curzons, A. D.; Jiménez -González, C.; Duncan, A. L.; Constable, D. J. C.; Cunningham, V. L. Fast life cycle assessment of synthetic chemistry (FLASC TM ) tool. Int. J. Life Cycle Assess. 2007, 12 (4), 272 Œ280. (66) Russell, D. A. M.; Shiang, D. L. Thi nking about more sustainable products: Using an efficient tool for sustainability education, innovation, and project management to encourage sustainability thinking in a multinational corporation. ACS Sustain. Chem. Eng. 2013, 1 (1), 2 Œ7. (67) Asatkar, A. K.; Bedi, A.; Zade, S. S. Metallo -organic Conjugated Systems for Organic Electronics. Isr. J. Chem. 2014, 54 (5Œ6), 467 Œ495. (68) Lizin, S.; Van Passel, S.; De Schepper, E.; Maes, W.; Lutsen, L.; Manca, J.; Vanderzande, D. Life cycle analyses of organic photovoltaics: a review. Energy Environ. Sci. 2013, 6 (11), 3136. (69) Zimmermann, Y. -S.; Schäffer, A.; Hugi, C.; Fent, K.; Corvini, P. F. -X.; Lenz, M. Organic photovoltaics: potential fate and effects in the environment. Environ. Int. 2012, 49 (SEPTEMBER ), 128 Œ140. (70) Wohrle, D.; Meissner, D. Organic Solar Cells. Adv. Mater. 1991 , 3, 129Œ138. (71) Williams, G.; Sutty, S.; Klenkler, R.; Aziz, H. Renewed interest in metal phthalocyanine donors for small molecule organic solar cells. Sol. Energy Mater. S ol. Cells 2014, 124, 217Œ226. (72) Peumans, P.; Forrest, S. R. Very -high -efficiency double -heterostructure copper phthalocyanine/C60 photovoltaic cells. Appl. Phys. Lett. 2001, 79 (1), 126 Œ128. 96 (73) J&K Scientific https://www.jk -scientific.com/ (accessed Apr 15, 2018). (74) Sigma -Aldrich (Millipore Sigma) https://www.sigmaaldrich.com/ (accessed Apr 15, 2018). (75) Swiss Centre For Life Cycle Inventories. Ecoinvent Database 3.3. Ecoinvent Centre . 2016. (76) Long Trail Sustainability. DATASMART LCI Package (US -EI SimaPro® Library) http://ltsexperts.com/services/software/datasmart -life -cycle -inventory/ (accessed Apr 15, 2018). (77) Jiménez -González, C.; Kim, S.; Overcash, M. R. Methodology for developing gate -to-gate Life cycle inventory information. Int. J. Life Cycle Assess. 2000 , 5 (3), 153 Œ159. (78) Geisler, G.; Hofstetter, T. B.; Hungerbühler, K. Production of fine and speciality chemicals: procedure for the estimation of LCIs. Int. J. Life Cycle Assess. 2004, 9 (2), 101Œ113. (79) Frischknecht, R.; Jungbluth, N.; Althaus, H.; Doka, G.; Dones, R.; Heck, T.; Hellweg, S.; Hischier, R.; Nemecek, T.; Rebitzer, G.; et al. Overview and Methodology ; Dübendorf, 2007. (80) Bare, J. TRACI 2.0: the tool for the reduction and assessment of chemical and othe r environmental impacts 2.0. Clean Technol. Environ. Policy 2011 , 13 (5), 687 Œ696. (81) Frischknecht, R.; Jungbluth, N.; Althaus, H.; Bauer, C.; Doka, G.; Dones, R.; Hellweg, S.; Humbert, S.; Köllner, T.; Loerincik, Y.; et al. Implementation of life cycle impact assessment methods ; Dübendorf, 2010. (82) Berger, M.; van der Ent, R.; Eisner, S.; Bach, V.; Finkbeiner, M. Water Accounting and Vulnerability Evaluation (WAVE): Considering Atmospheric Evaporation Recycling and the Risk of Freshwater Depletion in Water Footprinting. Environ. Sci. Technol. 2014 , 48 (8), 4521 Œ4528. (83) Globally Harmonized System of Classification and Labelling of Chemicals (GHS) https://www.unece.org/trans/danger/publi/ghs/ghs_welcome_e.html (accessed Apr 15, 2018). (84) Alder, C. M.; Hayler, J. D.; Henderson, R.; Redman, A.; Shukla, L.; Shuster, L. E.; Sneddon, H. F. Updating and further expanding GSK™s solvent sustainability guide. Green Chem. 2016, 4 (8), 1166 Œ1169. (85) Tomoda, H.; Saito, S.; Shiraishi, S. Synthesis o f metallophthalocyanines from phthalonitrile with strong organic bases. Chem. Lett. 1983, No. 3, 313 Œ316. (86) Boethling, R. S.; Sommer, E.; DiFiore, D. Designing Small Molecules for Biodegradability. Chem. Rev. 2007, 107 (6), 2207 Œ2227. 97 (87) Sheldon, R. A. The E factor 25 years on: the rise of green chemistry and sustainability. Green Chem. 2017, 19 (1), 18 Œ43. (88) Sheldon, R. A. The E Factor: fifteen years on. Green Chem. 2007, 9 (12), 1273. (89) Sijbesma, F.; Bock, K.; Clamadieu, J. -P. Social Life Cycle Metrics for Chemical Products: A guideline by the chemical sector to assess and report on the social impact of chemical products, based on a life cycle approach ; Geneva, 2016. (90) Rosenbaum, R. K .; Bachmann, T. M.; Gold, L. S.; Huijbregts, M. A. J.; Jolliet, O.; Juraske, R.; Koehler, A.; Larsen, H. F.; MacLeod, M.; Margni, M.; et al. USEtox - The UNEP -SETAC toxicity model: Recommended characterisation factors for human toxicity and freshwater ecot oxicity in life cycle impact assessment. Int. J. Life Cycle Assess. 2008 , 13 (7), 532 Œ546. (91) Toropova, A. P.; Toropov, A. a.; Benfenati, E.; Gini, G. QSAR modelling toxicity toward rats of inorganic substances by means of CORAL. Cent. Eur. J. Chem. 2010, 9 (1), 75 Œ85. (92) Melnikov, F.; Kostal, J.; Voutchkova -Kostal, A.; Zimmerman, J. B.; T. Anastas, P. Assessment of predictive models for estimating the acute aquatic toxicity of organic chemicals. Green Chem. 2016, 18 (16), 4432 Œ4445. (93) Burczyk, A. ; Loupy, A.; Bogdal, D.; Petit, A. Improvement in the synthesis of metallophthalocyanines using microwave irradiation. Tetrahedron 2005, 61 (1), 179 Œ188. (94) Sharma, R. K.; Sharma, C.; Sidhwani, I. T. Solventless and One -Pot Synthesis of Cu(II) Phthalocy anine Complex: A Green Chemistry Experiment. J. Chem. Educ. 2011, 88 (1), 86Œ87. (95) Lee, E.; Traverse, C. J.; Young, M.; Lunt, R. R.; Anctil, A. Evaluation of ClAlPc synthesis methods for transparent organic photovoltaic. In 2016 IEEE 43rd Photovoltaic Specialists Conference (PVSC) ; IEEE, 2016; pp 1649 Œ1652. (96) Japan Approves Ban on Polychlorinated Naphthalene, Pentachlorophenol http://industries.ul.com/news/japan -approves -ban -on-polychlorinated -naphthalene -pentachlorophenol (accessed Dec 22, 2017). (97) Environment and Climate Change Canada. Ecological screening assessment of chlorinated naphthalenes https://www.ec.gc.ca/ese -ees/default.asp?lang=En&n=835522FE -1#a12 (accessed Apr 15, 2018). (98) Prat, D.; Wells, A.; Hayler, J.; Sneddon, H. F.; McElro y, C. R.; Abou -Shehada, S.; Dunn, P. J. CHEM21 selection guide of classical - and less classical - solvents. Green Chem. 2015. (99) Owen, J. E.; Kenney, M. E. Phthalocyaninoaluminum Compounds. Inorg. Chem. 1962, 1 (2), 331 Œ333. 98 (100) Althaus, H. -J.; Chud acoff, M.; Hischier, R.; Jungbluth, N.; Osses, M.; Primas, A. Life cycle inventories of chemicals. Ecoinvent report 2007. (101) ECHA Registered Substances Database https://echa.europa.eu/web/guest/information -on-chemicals/registered -substances (accessed D ec 22, 2017). (102) Sharma, R. K.; Sharma, C.; Sidhwani, I. T. Solventless and one -pot synthesis of Cu(II) phthalocyanine complex: A green chemistry experiment. J. Chem. Educ. 2011, 88 (1), 86 Œ87. (103) metal -free and metallophthalocyanines containing four pyridyl groups. Transit. Met. Chem. 2007, 32, 851Œ856. (104) Tremblay, J. -F. Microwaving by the ton: Could indus trial -scale microwaves transform chemical manufacturing? Chemical & Engineering News . September 2016, pp 24 Œ25. (105) Microwave Chemical CO., Ltd. http://mwcc.jp/en/ (accessed Apr 11, 2018). (106) Emissions & Generation Resource Integrated Database https ://www.epa.gov/energy/emissions -generation -resource -integrated -database -egrid. (107) Wernet, G.; Mutel, C.; Hellweg, S.; Hungerbühler, K. The Environmental Importance of Energy Use in Chemical Production. J. Ind. Ecol. 2011, 15 (1), 96 Œ107. (108) EIA. So urce of U.S. electricity generation, 2017 https://www.eia.gov/energyexplained/index.php?page=electricity_in_the_united_states (accessed May 8, 2018). (109) EIA. Energy consumption of building sector https://www.eia.gov/tools/faqs/faq.php?id=86&t=1 (access ed May 8, 2018). (110) Chae, Y. T.; Kim, J.; Park, H.; Shin, B. Building energy performance evaluation of building integrated photovoltaic (BIPV) window with semi -transparent solar cells. Appl. Energy 2014, 129, 217Œ227. (111) Ng, P. K.; Mithraratne, N.; Kua, H. W. Energy analy sis of semi -transparent BIPV in Singapore buildings. Energy Build. 2013, 66, 274Œ281. (112) Leite Didoné, E.; Wagner, A. Semi -transparent PV windows: A study for office buildings in Brazil. Energy Build. 2013, 67, 136Œ142. (113) Olivieri, L.; Caamaño -Mar tín, E.; Moralejo -Vázquez, F. J.; Martín -Chivelet, N.; Olivieri, F.; Neila -Gonzalez, F. J. Energy saving potential of semi -transparent photovoltaic elements for building integration. Energy 2014, 76, 572Œ583. (114) EIA. Commercial Buildings Energy Consump tion Survey https://www.eia.gov/consumption/commercial/ (accessed May 8, 2018). 99 (115) Olivieri, L.; Caamaño -Martin, E.; Olivieri, F.; Neila, J. Integral energy performance characterization of semi -transparent photovoltaic elements for building integration under real operation conditions. Energy Build. 2014, 68 (PARTA), 280 Œ291. (116) Jayathissa, P.; Jansen, M.; Heeren, N.; Nagy, Z.; Schlueter, A. Life cycle assessment of dynamic building integrated photovoltaics. Sol. Energy Mater. Sol. Cells 2016 , 156, 75Œ82. (117) Patil, B. R.; Mirsafaei, M.; Cielecki, P. P.; Cauduro, A. L. F.; Fiutowski, J.; Rubahn, H. G.; Madsen, M. ITO with embedded silver grids as transparent conductive electrodes for large area organic solar cells. Nanotechnology 2017 , 28 (40), 40 5303. (118) Hermenau, M.; Riede, M.; Leo, K. Degradation of Small -Molecule -Based OPV. In Stability and Degradation of Organic and Polymer Solar Cells ; John Wiley & Sons, Ltd: Chichester, UK, 2012; pp 109 Œ142. (119) Mendon, V.; Taylor, Z.; Rao, S.; Xie, Y . 2015 IECC Determination of Energy Savings: Preliminary Technical Analysis ; 2014. (120) ICC. 2015 International Energy Conservation Code https://codes.iccsafe.org/content/IECC2015/chapter -3-ce-general -requirements (accessed Nov 12, 2018). (121) Marion, W.; Wilcox, S. Solar radiation data manual for buildings ; Golden, CO (United States), 1995. (122) EIA. Energy Explained: what is a degree day? https://www.eia.gov/energyexplained/index.php?page=about_degree_days (accessed Nov 12, 2018). (123) DOE. Energy Plus https://energyplus.net/. (124) Deru, M.; Field, K.; Studer, D.; Benne, K.; Griffith, B.; Torcellini, P.; Liu, B.; Halverson, M.; Winiarski, D.; Rosenberg, M.; et al. U.S. Department of Energy Commercial Reference Building Models of the National Build ing Stock ; Golden, CO (United States), 2011. (125) EnergyPlus https://energyplus.net/weather. (126) Markvart, T.; CastaÑer, L. Principles of Solar Cell Operation. In Practical Handbook of Photovoltaics ; Elsevier, 2003; pp 71 Œ93. (127) U.S. Census Bureau . 2017 American Community Survey 1 -Year Estimates ; 2017. (128) ASHRAE. Standard 90.1 -2016, Energy Standard for Buildings Except Low -Rise Residential Buildings ; 2016. (129) World™s population increasingly urban with more than half living in urban areas 100 http://www.un.org/en/development/desa/news/population/world -urbanization -prospects -2014.html. (130) EPA. Reducing Urban Heat Islands: Compendium of Strategies Urban Heat Island Basics ; 2014. (131) Crawley, D. B. Estimating the impacts of climate change and urbanization on building performance. J. Build. Perform. Simul. 2008, 1 (2), 91 Œ115. (132) Bueno, B.; Norford, L.; Hidalgo, J.; Pigeon, G. The urban weather generator. J. Build. Perform. Simul. 2013, 6 (4), 269 Œ281. (133) Yang, J. H. The curious case of urban heat island: a systems analysis, Massachusetts Institute of Technology, 2016. (134) U.S. Environmental Protection Agency. Heat island effect https://www.epa.gov/heat -islands. (135) Memon, R. A.; Leung, D. Y. C.; Liu, C. -H. An investigation of urba n heat island intensity (UHII) as an indicator of urban heating. Atmos. Res. 2009, 94 (3), 491 Œ500. (136) Barron -Gafford, G. A.; Minor, R. L.; Allen, N. A.; Cronin, A. D.; Brooks, A. E.; Pavao -Zuckerman, M. A. The photovoltaic heat island effect: Larger solar power plants increase local temperatures. Sci. Rep. 2016, 6 (September), 1 Œ7. (137) Van Aken, K.; Str ekowski, L.; Patiny, L. EcoScale, a semi -quantitative tool to select an organic preparation based on economical and ecological parameters. Beilstein J. Org. Chem. 2006, 2 (1), 3. (138) Ribeiro, M. G. T. C.; Costa, D. A.; Machado, A. A. S. C. fiGreen Starfl: a holistic Green Chemistry metric for evaluation of teaching laboratory experiments. Green Chem. Lett. Rev. 2010, 3 (2), 149 Œ159. (139) Mercer, S. M.; Andraos, J.; Jessop, P. G. Choosing the Greenest Synthesis: A Multivariate Metric Green Chemistry Exerc ise. J. Chem. Educ. 2012, 89 (2), 215 Œ220. (140) Gonzalez, M. A.; Smith, R. L. A Methodology to Evaluate Process Sustainability. Environ. Prog. 2003, 22 (4), 269 Œ276. (141) Ruiz -Mercado, G. J.; Carvalho, A.; Cabezas, H. Using Green Chemistry and Engineer ing Principles To Design, Assess, and Retrofit Chemical Processes for Sustainability. ACS Sustain. Chem. Eng. 2016 , 4 (11), 6208 Œ6221. (142) McElroy, C. R.; Constantinou, A.; Jones, L. C.; Summerton, L.; Clark, J. H. Towards a holistic approach to metrics for the 21st century pharmaceutical industry. Green Chem. 2015, 17 (5), 3111 Œ3121. (143) Koller, G.; Fischer, U.; Hungerbühler, K. Assessing Safety, Health, and Environmental 101 Impact Early during Process Development. Ind. Eng. Chem. Res. 2000, 39 (4), 960 Œ972. (144) Eissen, M.; Metzger, J. O. Environmental performance metrics for daily use in synthetic chemistry. Chem. - A Eur. J. 2002, 8 (16), 3580 Œ3585. (145) Andraos, J. Inclusion of environmental impact parameters in radial pentagon material efficiency metrics analysis: Using benign indices as a step towards a complete assessment of fiGreennessfl for Chemical Reactions and Synthesis Plans. Org. Process Res. Dev. 2012, 16 (9), 1482 Œ1506. (146) Andraos, J. Safety/Hazard Indices: Completion of a unified suite of metrics for the assessment of figreennessfl for chemical reactions and synthesis plans. Org. Process Res. Dev. 2013, 17 (2), 175 Œ192. (147) Eckelman, M. J. Life Cycle Inherent Toxicity: A Novel LCA -based Algorithm for Evaluating Chemical Synthesis Pathways. Green Chem. 2016 . (148) Smith, R. L.; Ruiz -Mercado, G. J.; Meyer, D. E.; Gonzalez, M. A.; Abraham, J. P.; Barrett, W. M.; Randall, P. M. Coupling Compute r-Aided Process Simulation and Estimations of Emissions and Land Use for Rapid Life Cycle Inventory Modeling. ACS Sustain. Chem. Eng. 2017 , 5 (5), 3786 Œ3794. (149) DeVierno Kreuder, A.; House -Knight, T.; Whitford, J.; Ponnusamy, E.; Miller, P.; Jesse, N.; Rodenborn, R.; Sayag, S.; Gebel, M.; Aped, I.; et al. A Method for Assessing Greener Alternatives between Chemical Products Following the 12 Principles of Green Chemistry. ACS Sustain. Chem. Eng. 2017, 5 (4), 2927 Œ2935. (150) Song, R.; Keller, A. A.; Suh , S. Rapid Life -Cycle Impact Screening Using Artificial Neural Networks. Environ. Sci. Technol. 2017, 51 (18), 10777 Œ10785. (151) Koster, K.; Cohen, M. Practical Approaches to Sustainability: iSUSTAIN® Tool for Green Chemistry Case Study. In Treatise on S ustainability Science and Engineering ; Springer Netherlands: Dordrecht, 2013; pp 81 Œ108. (152) Blümer, G. -P.; Collin, G.; Höke, H. Tar and Pitch. In Ullmann™s Encyclopedia of Industrial Chemistry ; Wiley -VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2011. (153) Hiers, G. S.; Hager, F. D. Anisole. Org. Synth. 1929 , 9, 12. (154) Moon, B. S.; Choi, H. Y.; Koh, H. Y.; Chi, D. Y. The use of sodium chlorate/hydrochloric acid mixtures as a novel and selective chlorination agent. Bull. Korean Chem. Soc. 2011, 32 (2), 472 Œ476. (155) Mettu, A. New synthesis for production of caprolactam by Beckmann rearrangement of chyclohexanone over differetn incorporated molecular sieves and oxide catalysts, RWTH Aachen University, 2009. 102 (156) Oediger, H.; Möller, F.; Eiter, K . Bicyclic Amidines as Reagents in Organic Syntheses. Synthesis (Stuttg). 1972, 1972 (11), 591 Œ598. (157) Bureau of Transportation Statistics and U.S. Census Bureau. Commodity flow survey: Hazardous Materials ; 2012. (158) Moseley, J. D.; Kappe, C. O. A critical assessment of the greenness and energy efficiency of microwave -assisted organic synthesis. Green Chem. 2011, 13 (4), 794. (159) Godwin, D. R.; Lawton, S. J.; Moseley, J. D.; Welham, M. J.; Weston, N. P. Energy Effic iency of Conventionally -Heated Pilot Plant Reactors Compared with Microwave Reactors. Energy & Fuels 2010, 24 (10), 5446 Œ5453.