Photodegradation in a stress and response framework: poly ......topic of photovoltaics (PV) lifetime...

Photodegradation in a stress andresponse framework poly(methylmethacrylate) for solar mirrors and lens

Myles P MurrayLaura S BruckmanRoger H French

Downloaded From httpspiedigitallibraryorg on 01292013 Terms of Use httpspiedlorgterms

Photodegradation in a stress and response frameworkpoly(methyl methacrylate) for solar mirrors and lens

Myles P Murray Laura S Bruckman and Roger H FrenchCase Western Reserve University Solar Durability and Lifetime Extension Center and

Materials Science Department Cleveland Ohio 44106rogerfrenchcaseedu

Abstract In the development of materials for enhanced photovoltaic (PV) performance it iscritical to have quantitative knowledge of both their initial performance and their performanceover the required 25-year warranted lifetime of the PV system Lifetime and degradation sciencebased on an environmental stress and response framework is being developed to link the inten-sity and net stress to which materials components and systems are exposed to the responsesobserved and their subsequent degradation and damage accumulation over the lifetime Inducedabsorbance to dose (IAD) a metric developed for solar radiation durability studies of solar andenvironmentally exposed materials is defined as the rate of photodarkening or photobleaching ofa material as a function of radiation dose Quantitative degradation rates like IAD determinedover a wide range of stress intensities and net stresses have the potential to predict degradationfailure and power loss rates in photovoltaic systems over time caused by damage accumulationTwo grades of poly(methyl methacrylate) were exposed and evaluated in two cases of high-intensity ultraviolet exposures A three- to six-fold increase in photodarkening was observedfor one acrylic formulation when exposed to UVA-340 light when compared with concentratedxenon-arc exposure The other more highly stabilized acrylic formulation showed up to threetimes more photodarkening in the same exposure copy 2012 Society of Photo-Optical InstrumentationEngineers (SPIE) [DOI 1011171JPE2022004]

Keywords photodegradation acrylic poly(methyl methacrylate) photovoltaics degradation

Paper 12020SS received Mar 21 2012 revised manuscript received Jul 26 2012 accepted forpublication Oct 1 2012 published online Nov 19 2012

1 Introduction

A recent US Department of Energyworkshop on Science for EnergyTechnologies1 identified thetopic of photovoltaics (PV) lifetime and degradation science (LampDS)2ndash4 as a critical scientificchallenge for robust adoption of PV The 25-year lifetime performance of PV requires a betterunderstanding of the degradation mechanisms in PV materials components and systems Bydeveloping metrics metrology and tools to quantify compare and cross-correlate the responseof PV systems components and materials to a variety of stressors such as ultraviolet (UV) radia-tion humidity and temperature variation for both accelerated and outdoor testing we can linkstresses to observed responses in a stress-response framework and determine quantitative rates ofdegradation LampDS requires the development of quantitative degradation mechanisms and ratesfor degradation and failure modes so as to enable quantitative lifetime projections

Lifetime and degradation science (LampDS) based on a stress and response [RethSTHORN] frameworkis being developed that links the intensity and net stress to which materials components andsystems are exposed to the responses observed and the degradation and damage accumulationover lifetime This RethSTHORN framework can encompass multifactor and cyclic environmental stres-sors including solar irradiance temperature and humidity which can cause degradation overtime The RethSTHORN framework allows for the cross-correlation of the materials response and degra-dation rates with numerous applied stress intensities and net stresses and allows determination ofthe effects of stressors even at accelerated intensities with applicability to multiple service con-ditions At the same time an RethSTHORN framework permits incorporation of data determined over

0091-32862012$2500 copy 2012 SPIE

Journal of Photonics for Energy 022004-1 Vol 2 2012


many stress conditions and corresponding to many different responses and its cross-correlationto produce a more complete picture of a materialrsquos or systemrsquos stress and response when com-pared to the traditional acceleration factors approach which is always in reference to one set ofservice or application conditions The RethSTHORN framework requires the use of quantitative metrics tobe used such as induced absorbance to dose (IAD) which is defined as the rate of photodarkeningor photobleaching of a material as a function of radiation dose56 In the RethSTHORN framework allof the available responses of a material are integrated to provide a comprehensive scientificunderstanding of the material47

Reliability engineering is thought to fall into three main statistical regimes infant mortalityrandom failures and failure at the end of life8 (Fig 1) Initial performance qualification testingrepresented by UL 17039 IEC 6210810 IEC 6121511 and IEC 6164612 allows for reasonableconfidence intervals when mitigating infant mortality and can sometimes be useful in predictingrandom failures in PV modules which are failures that occur imminently or randomly afterdeployment due to process and manufacturing defects1314 Lifetime performance howeverrequires knowledge of the power degradation rate of a PV system or module a quantity that iscurrently only quantified after deployment has already occurred It is thought that by performinghighly accelerated stress tests degradation modes of materials can be elucidated and withconfirmation from outdoor testing lifetime performance can be predicted

Mirror augmented photovoltaic systems utilize mirrors to couple more light onto a PV mod-ule and its absorber By increasing the light absorbed they have the potential to provide a lowerlevelized cost of electricity (LCOE)15 Low concentration photovoltaic (LCPV) systems wheresolar irradiance is concentrated by a factor of 1 to 10 present opportunities for cost benefits whencompared with traditional technologies In these systems electrical output per unit area of activematerials increases nearly linearly with concentration factor thereby reducing the cost of activematerials per watt by up to a factor of 10 while still having the opportunity to utilize traditional 1sun PV technologies At the same time increased solar irradiance from LCPV amplifies thestress intensity and possibly the degradation rates of the systems components and materialsAdditionally there is extensive interest in high concentrating photovoltaics applications for useof Fresnel lenses fabricated from durable poly(methyl methacrylate) (PMMA) grades16

Polymers of many types are vulnerable to UV degradation because the highest energy UVradiation observed in the solar spectrum has energies that surpass all but the strongest carbonbonds In order to protect these polymers various formulations of UVabsorbers hindered aminelight stabilizers (HALs) and radical scavengers have been employed to stabilize materialsexposed to UV degradation17 A common class of UV absorbers are hydroxyphenyl benzotri-azole compounds such as Tinuvinreg made by BASF1819 which are usually used at concentrationsbetween 01 and 1 wt 20 and protect the polymer matrix by being sacrificially degraded by theUV light18 The degradation of numerous grades and formulations of PMMA acrylic polymershas been studied under a wide variety of UV light sources including mercury vapor lights emit-ting at 2537 nm21ndash25 xenon arc lamps filtered to match AM 1526 and 60Co emitted gammaradiation21 Additionally studies of the wavelength dependence of the photodegradation

Fig 1 Reliability engineering curve showing the three main statistical regimes infant mortalityrandom failures and failure at the end of life

Murray Bruckman and French Photodegradation in a stress and response framework



have been performed using glass filters26 monochromators2728 and with different light sourcesemitting at 300 nm and 350 nm22 Indoor exposures have been performed under vacuum and inair with mass loss metrology and gaseous degradation products being identified by mass spec-troscopy222425 Production of degradation products and viscosity averaged molecular weightdistributions have been used to determine the quantum efficiency of UV light in degradingPMMA2124252729 Others have performed outdoor exposures with flat plate exposures263031

and with accelerated tracking exposures27

Optical characterization of degradation in acrylics has been done using colorimetry2631 opti-cal density2829 UV spectroscopy1624ndash3032ndash35 electron spin resonance spectroscopy21 and Fouriertransform infrared spectroscopy1628303235 Surface degradation of acrylics has been studiedusing atomic force microscopy16 scanning electron microscopy30 and contact angle measure-ments16 Mechanical testing has been used to characterize the effects of UV degradation onPMMA mechanical properties3134 Other studies have been done to analyze the thermaldecomposition and depolymerization of PMMA associated with mass loss and increasedphotodegradation36

Degradation of PMMA by ultraviolet light is thought to occur when high energy incidentradiation initiates radicalization of an ester side group leading to β-scission of the polymerbackbone and decomposition of the radical end to products including H2 CO CO2 CH2OH

and HCOOCH325 The loss of the ester group creates a greater free volume in the PMMA

increasing likelihood of creep and dimensional instability37 Chain scission that occurs fromexposure to solar radiation is predominantly caused by UV radiation with wavelengths shorterthan 320 nm2728 This process is accompanied by an increase in yellowing in many experimentsThe color centers responsible are identified as free radicals trapped with the bulk material untilthey become neutralized by oxygen or other radiation38

Typically UV exposure of materials can be performed with a xenon arc light source whichcan be filtered to closely match the shape of the solar spectrum for air mass (AM) 15 a standardrepresenting real-world exposure at 482 deg latitude39 AM 15 exposures can be performed atmultiple levels of irradiance by using concentrating optics to focus the light beam This enablesproper weathering acceleration because spectral matching between outdoor conditions andindoor exposures excludes degradation induced by radiation outside of typical environmentalconditions However due to their lower operating costs fluorescent exposure of materials usingUVB-313 UVA-340 or UVA-351 lamps are often used as a less expensive way to provideinsights into material performance over time UVA-340 lamps radiate light that closely matchesthe AM 15 spectra in the region from 280 to 360 nm when the spectra radiance peaks at 340 nmwith a value of 030 W∕m2∕nm

As a case study to demonstrate the LampDS approach to lifetime and degradation science wereport here a study on photodegradation in two acrylic PMMA grades under two different irra-diation sources A Q-Lab QUV accelerated weathering tester40 outfitted with UVA-340 lampswas used to expose samples to UV radiation this method for exposure was compared with aNewport solar simulator equipped with a concentrator The ASTM G154 Cycle 4 withoutthe condensation step QUV exposure was performed with spectral irradiance peaking at155 W∕m2∕nm at 340 nm while the xenon arc exposure was performed with full spectral irra-diance (integrated over the AM 15 spectrum) controlled at 504 kW∕m2 The temperature insidethe QUV tester was kept at 70degC with a sample temperature of approximately 67degC while tem-perature inside the Newport enclosure was uncontrolled producing a typical sample temperatureof approximately 50degC41 Different formulations of PMMAwere exposed to these single-factoraccelerated exposures Average spectral IADs and yellowing indices (YI) for the exposures weredetermined and correlations between exposures are discussed

2 Stress-Response [RS] Framework for Solar Degradation

Stressors that impact PV materials and components can be characterized in terms of instanta-neous stress level (σ) and net stress or integrated stress (S) which is the instantaneous stresslevel integrated over the length of time the stress was applied [Eq (1)] Changing the instanta-neous stress level may change the materialrsquos response characteristics therefore stressors must bequantified in terms of both instantaneous stress level and integrated stress [Eq (2)]




S frac14Z

σdt (1)

XS frac14

Zethσi otimes σj otimes σnTHORNdt (2)

where σi and σj are different stress types and levels and otimes represents the convolution of thosestresses

A materialrsquos response (R) to both instantaneous stress level and integrated stress maycorrespond to a change in the optical properties of a material a loss of mechanical strengthor any measurable change in properties arising due to stressors applied over time The generalrelationship between stress and response is a function of that stress [Eq (3)]

R frac14 fethσSTHORN frac14Z

fethσTHORNσdt (3)

Single factor accelerated testing is often performed on materials to determine their durabilityto a particular environmental stress By varying the stress intensity it is possible to show therange of conditions that do not open new degradation pathways An expansion of thismethodology to multiple stresses gives [Eq (4)]

RethSi Sj SnTHORN frac14Z

fethσi σj σnTHORNethσi otimes σj otimes σnTHORN (4)

where Si Sj Sn are different stresses that cause responses The benefit of such a framework isthat stresses can be separated and accounted for in this methodology However single stressexposures introduce hazardous assumptions because some stresses have synergistic effectsmeaning that a correlation function between stress and response in real-world environmentscan depend on all stresses applied to the system A multifactor test which includes multiplestressors can lead to a better understanding of the synergistic effects of stress in a real-world environment Response is therefore a function of the convolution of multiple stressorsat their service-use conditions The convolution of these stresses implies that they can havesynergistic effects Therefore the correlation function which describes the impact of multiplestressors in environmental conditions while unknown depends on all stresses applied to thesystem If an accelerated exposure can be shown to induce the same response as observedin environmental conditions a correlation function can be used to predict the response observedin a system exposed to environmental conditions without waiting for real-time testing results

3 Methods

31 PMMA Formulations

Two formulations of PMMA used in the present study are multipurpose (MP) acrylic and UVtransparent (UVT) acrylic samples Each contains different amounts of the same UV stabilizerThese acrylic samples were from Replex Plastics Inc42 MP grade acrylic is used in the securityand transportation industry as a substrate for mirrors and housings while UVT is used for day-lighting43 applications that require full spectrum light The thicknesses of the acrylic sampleswere approximately 3 mm

32 Simulated Solar Exposures

Exposures were performed with a Q-Labs QUVAccelerated Weathering Tester (Model QUVSpray with Solar Eye Irradiance Control)40 and with a Newport 16 kW diverging beam solarsimulator (Model 92190) with a 13times irradiance concentrator (Model SP81030-DIV) The twodifferent stress conditions are compared in Table 1 By sequential measurements of optical




properties after different steps of exposure dose (or exposure doses) it is possible to determinethe relationship between solar irradiance (the stress) and induced degradation rates of opticalproperties (the response) of these materials along with the form of the dose dependence ofthe response (linear exponential sub- or supra-linear) This 18-day exposure is equivalentto radiant exposure of 86 years of tracked exposure in Phoenix Arizona44

The QUV UVA-340 exposures had eight of each type of acrylic samples so as to providesufficient statistics A representative sample was removed after each dose step and the baselinesample was never exposed The first dose step was 284 h and each subsequent dose step was220 h There was only one sample for each type of acrylic exposed in the Newport solar simu-lator at 504 kW∕m2 Each dose step was 72 h The baseline abscm and YI measurements wereperformed on the acrylic samples prior to exposures

321 QUV accelerated weathering tester

The QUVaccelerated weathering tester uses customized fluorescent lamps to expose samples tohigh doses of UV radiation which simulates different types of damaging environmental stressconditions The exposures were performed with UVA-340 lamps which emit radiation between280 and 400 nm This closely matches the AM 15 spectrum at wavelengths shorter than 360 nmwhere much of the damaging radiation exists in the solar spectrum The QUV was run usingASTM G154 Cycle 4 (155 W∕m2∕nm at 340 nm at 70degC) without the condensation step for21 days41 By setting the QUV peak intensity to 155 W∕m2∕nm at 340 nm a higher level ofdamage can be induced making this a single-factor accelerated test compared to outdoor weath-ering This level of UV radiation is 52 times higher than the intensity of AM 15 at 340 nmWhile the integration of the QUV spectra gives a much lower total dose the ratio of light in thedamaging range from 280 to 360 nm is much higher Only 192 of AM 15 radiation falls intothe UVA-340 exposure wavelengths By correcting the TMY3 data for this factor the 944 hexposure is equivalent to 12 years of tracked outdoor exposure in Arizona44 This was calculatedwith Eqs (5) and (6) where Eeλ is spectral irradiance for each source (W∕m2∕nm) λ is wave-length in nm Ee is irradiance (W∕m2) and He is radiant exposure to wavelengths between280 nm and 360 nm (J∕m2)45

Ee frac14Z

360

280

Eeλdλ (5)

Table 1 A comparison of the stress conditions

StressorsNewport solarsimulator (50times)

QUV (155 W∕m2∕nmat 340 nm)

Total irradiance Full spectrum 280 to 4000 nm 504 kW∕m2 845 W∕m2

UVA-340 280 to 360 nm 576 W∕m2 6065 W∕m2

TUV 280 to 400 nm 1570 W∕m2 845 W∕m2

Dose step 1306 GJ∕m2 AM15 62 and 48 MJ∕m2 UVA-340

Total dose Full spectrum 280 to 4000 nm 785 GJ∕m2 287 MJ∕m2

UVA-340 280 to 360 nm 151 GJ∕m2 206 MJ∕m2

TUV 280 to 400 nm 244 GJ∕m2 287 MJ∕m2

Time (days) 18 39

Sample temperature (degC) 50 67




He frac14Z

Eedt frac14Z

t

0

Z360

280

Eeλdλ (6)

322 Newport diverging solar simulator

The Newport xenon arc solar simulator46 was used to irradiate material samples with spectralmatched AM 15 radiation39 using a Newport 13x high flux concentrator accessory with the solarsimulator irradiance levels of 504 kW∕m2 were achieved making this a single-factor acceler-ated test compared to both UVA-340 and outdoor exposure Continuous power monitoring with aNewport power meter (Model 1918-R) was used to monitor the irradiance making it possible toquantify the irradiance and dose to which a sample was exposed Integrated dose was calculatedfor total dose and UVA-340 [Eqs (5) and (6)] Compared with outdoor exposures in Arizonausing typical metrological year data (TMY3) supplied by NREL44 which are about 91 GJ∕m2

per year this 18-day test exposed samples to doses equivalent to approximately 85 years

33 Evaluations

331 Cary 6000i with DRA-1800

An Agilent Cary 6000i spectrometer with a diffuse reflectance accessory-1800 (DRA) and anInGaAs detector was used to determine the Abscm for each acrylic sample The spectra wereacquired from 180 to 1800 nm every 040 nm with a scan rate of 4800 nm∕min and a spectralbandwidth of 200 nm in the UV-Vis and 400 nm in the NIR regions

332 Hunterlab UltraScan Pro

The HunterLab UltraScan Pro was used to determine the YI of each of the acrylic samples Thisinstrument is a fast high-performance color measurement spectrophotometer with a spectralrange from 350 to 1050 nm with a 5 nm optical resolution

333 Temperature

The temperature for the samples in the Newport solar simulator and the QUVaccelerated weath-ering tester were determined with a FLIR infrared camera (Model T300)47

34 Metrics

341 Induced absorbance to dose

Induced absorbance to dose (IAD) is a measurement of the change at a particular wavelength ofthe bulk optical absorbance per centimeter base ten (Abscm) of a material and this analysisassumes uniform bulk absorption phenomena which may not be observed in certain samplesTherefore IAD analysis is assumed to be independent of thickness In these samples preliminaryevidence shows that photodegradation progresses from the light-exposed surface through thebulk of the material Therefore the IAD metric as used here represents the average of the opticalabsorbance per centimeter through the thickness of the sample IAD is a quantitative dose metricfor photodarkening and photobleaching on a per unit dose basis4849 The per unit dose basis canbe changed so that IAD can be calculated per full spectrum dose or per UVA-340 dose Incre-mental IAD is useful to track and identify transient phenomena which tend to stop after a certaindose (eg phenomena associated with curing or photobleaching of impurities) While averageIAD is useful to follow over large doses and long exposure times since the total dose exposedincreases the contribution due to transient phenomena are reduced in amplitude and continuousdegradation processes end up being the predominant observed process in the average IADIf either calculation of IAD reports the same values for successive dose steps then the materialis behaving linearly and if the same values of IAD are seen with two different exposures then




reciprocity in irradiance is being obeyed The average IAD [Eq (7)] and the incremental IAD[Eq (8)] are given by

AverageAbs

cmper

GJ

m2dose frac14 AbstethλTHORN∕cm minus AbsiethλTHORN∕cm

Doset(7)

IncrementalAbs

cmper

GJ

m2dose frac14 Absithorn1ethλTHORN∕cm minus AbsiethλTHORN∕cm

Doseithorn1 minus Dosei (8)

IAD values were determined with the Agilent Cary 6000i with a DRA-1800

342 Yellowness index

Yellowness index (YI) as defined in the ASTM E31350 is a colorimetric measure of yellowingBecause YI is measured over a broad wavelength range it is more sensitive than typical spectralmeasurements YI is calculated from the transmittance values for λ frac14 380 to 780 nm50 Yellow-ness indices are also useful because they are closely linked to reduced optical performance16

YI can be determined by Eq (9)

YI frac14 1000ethCxX minus CzZTHORNY

(9)

where X Y and Z are the CIE tristimulus values For D65∕10deg Cx frac14 13013 Cz frac14 1149850

YI was determined using the HunterLab UltrascanPro colorimeter Yellowing rates the rate ofchange in yellowness as a function of dose were determined by Eq (10)

RYI frac14dethYITHORNdethSTHORN (10)

4 Results

The absorbance per centimeter base 10 results of two different types of exposures on two for-mulations of acrylic PMMA UVT and MP are reported in Figs 2 and 3 respectively Baselinespectra of both formulations of acrylic show a fundamental absorption edge near 275 nm

Fig 2 (a) Abscm for ultraviolet transparent (UVT) grade poly(methyl methacrylate) (PMMA) forexposures in the QUV accelerated weathering tester to UVA-340 irradiance for baseline one tofour dose samples (b) Abscm for UVT grade PMMA for exposures in the Newport solar simulatorto AM 15 irradiance for baseline one to six dose samples




(region 1) however MP grade acrylic shows two strong absorption peaks centered near 298 nm(region 2) and 339 nm (region 2prime) due to the absorber package while both formulations aretransparent to wavelengths longer than 380 to 780 nm (region 3) but IAD are calculated at440 nm for region 3

41 IAD Results

The IAD values for the two exposures and PMMA grades of acrylic are summarized in Table 2for regions 1 2 2prime and 3 at equivalent dose steps UVT acrylic exposures in the QUV andNewport solar simulator show the UV degradation of unprotected PMMA acrylic (Fig 4)Because this material does not contain a significant UV absorber package UVA-340 dose ishighly effective at degrading this polymer Photodarkening is shown as a positive IAD curveand is observed at all wavelengths but increases near the fundamental absorption edge

MP acrylic exposures in QUVand Newport solar simulator show the UV degradation of UVformulated PMMA acrylic (Fig 5) This material which contains a significant UV absorberpackage is highly protected from UV degradation Photodarkening is again observed at thefundamental absorption edge shown as positive IAD values however both studies showsome initial photobleaching of UV stabilizers and slight photodarkening in the visible region

Fig 3 (a) Abscm for multipurpose (MP) grade poly(methyl methacrylate) (PMMA) for exposuresin the QUV accelerated weathering tester to UVA-340 irradiance for baseline one to four dosesamples (b) Abscm for MP grade PMMA for exposures in the Newport solar simulator to AM 15irradiance for baseline one to six dose sample

Table 2 Average IAD values for each region in the two stressors and two grades of PMMA acrylicfor equivalent dose steps Equivalent dose steps were 251 MJ∕m2 of UVA-340 dose for theQUV accelerated weathering tester and 206 MJ∕m2 of UVA-340 dose for the Newport solarsimulator IAD values were calculated with Eq (7) which were at step 4 and step 1 respectively

ExposuresPMMAgrade

Region 1(275 nm)

Region 2(298 nm)

Region 2prime(339 nm)

Region 3(440 nm)

QUV UVT 0019 00069 00029 000070

MP 00042 minus000042 minus00018 705 times Eminus5

Newport solar simulator UVT 00069 00021 000062 000011

MP 00036 000045 minus000033 18 e-5




42 Yellowness Indices

Yellowness indices are reported for samples exposed to UVA-340 radiation (Fig 6) The yellow-ing rates for UVTand MP are 0041 and 00035 1∕ethMJ∕m2THORN) of UVA-340 respectively Yellow-ing rates for Newport solar simulator became non-linear after two dose steps but yellowing ratesin the linear region for UVT and MP are 00068 and 00014 1∕ethMJ∕m2THORN respectively

43 Temperature

The temperature of the sample in the Newport solar simulator was approximately 50degC while thesample temperature in the QUV accelerated weathering tester was approximately 67degC

Fig 4 (a) Average induced absorbance to dose (IAD) of ultraviolet transparent (UVT) gradepoly(methyl methacrylate) (PMMA) after exposure in the QUV accelerated weathering tester toUVA-340 irradiance (b) Average IAD of UVT acrylic after exposure in the Newport solar simulatorto AM 15 irradiance

Fig 5 (a) Average induced absorbance to dose (IAD) of multipurpose (MP) grade poly(methylmethacrylate) (PMMA) after exposure in the QUV accelerated weathering tester to UVA-340irradiance (b) Average IAD of MP grade PMMA after exposure in the Newport solar simulator toAM 15 irradiance




5 Discussion

51 UVA-340 Dose Metric

In our quantification of exposure stresses UVA-340 dose was chosen as an irradiance dosemetric instead of the traditional TUV dose295152 UVA-340 dose is a useful metric becauseUVA-340 lamps have peak intensity at 340 nm and their spectral irradiance drops off quicklyat wavelengths longer than 360 nm UVA-340 dose includes the most damaging wavelengths oflight that cause degradation in the PMMA samples The wavelengths between 360 to 400 nm thatare included in the TUV metric have not been shown to cause significant degradation and thespectral match to the light source is not close Using UVA-340 dose for calculating response ratesis expected to show better correlation results of QUVaccelerated weathering tester and Newportsolar simulator exposures because TUV dose would inherently underestimate the amount ofdamaging short-wave UV light present in QUV exposure with UVA-340 bulbs (Table 1)

52 Comparison of Stressors

While photodegradation mechanisms are many complex and can be ill-defined photodarken-ing photobleaching and yellowing rates can be discussed without the complexities of explicitmechanistic insights Higher amounts of total degradation were observed in all samples exposedin the Newport solar simulator shown by the larger increases in optical absorbance for all sam-ples This was expected because of the fact that these samples saw much higher net stressesHowever when comparing responses of samples on a degradation per dose basis shown withIAD values QUVexposure showed a higher rate of degradation (Table 3) The QUVaccelerated

Fig 6 Yellowness index comparison of ultraviolet transmitting (UVT) and multipurpose (MP)grade poly(methyl methacrylate) after QUV accelerated weathering tester exposure to UVA-340irradiance

Table 3 The response ratios of UVT acrylic to MP acrylic for eachtype of stress The response ratios are determined from the IADvalues in Table 2 for equivalent dose steps between the two stressors

Stressor Region 1 (275) Region 3 (440 nm) YI

QUV 45 99 117

NSS 19 61 51




weathering tester is a 5x UV irradiance stress at 155 W∕m2∕nm at 340 nm of UVA-340 whilethe Newport solar simulator is a 50x full spectrum irradiance stress

The IAD curves for dose steps which are comparable in terms of net dose between the twoexposures (Fig 7) show that exposures in the QUV protocol have a larger IAD value in eachregion than an equivalent dose from the Newport solar simulator This may be due to the effect ofthe elevated chamber temperature as summarized in Table 1 In the UVT acrylic three timesmore photodarkening occurred near the fundamental absorption edge and six times more photo-darkening occurred from 350 to 500 nm In the MP acrylic equivalent photodarkening occurrednear the fundamental absorption edge and three times more photodarkening occurred from 375to 450 nm when comparing the two stressors However the stabilizer package shows a muchhigher level of photobleaching in the QUV exposure protocol which supports the assumptionthat the QUV protocol induces more degradation and that this effect may be due to the highersample temperature The sample temperatures in the QUV were approximately 67degC and in theNewport solar simulator were approximately 50degC

53 Comparison of PMMA Grades

IAD values for each of the two PMMA grades for each exposure show that the IAD values werehigher for UVT acrylic compared to MP acrylic in both stress conditions as seen in Table 2MP shows features in all three regions while UVT does not have distinct spectral features inregion 2 In all samples exposed there is significant photodarkening occurring in region 1 InQUVexposure of MP samples there is some initial photobleaching in regions 2 and 2prime howeversubsequent dose steps show photodarkening of the base polymer lowering this photobleachingand eventually leading to photodarkening in region 2 Region 3 is characterized by broad bandincrease in optical absorbance This response is decreased by a factor of nearly 10 with theaddition of the stabilizer package in MP grade acrylic

In addition to the effects of temperature another factor which may account for the increasedrate of photodegradation in the UVT sample in the QUV is that the sample did not screen any UVlight which passed through the sample Because samples in the QUV were mounted on alu-minum trays degrading UV light may have reflected passing through the sample a second timewhile in MP samples the UV light was absorbed by the UV stabilizer package and only lowerenergy light was reflected This may account for the differences in IAD rates at the fundamentalabsorption edge

The response ratios between the grades show that more degradation occurred in UVT inregion 1 and 3 for both stress exposures This relates to the fact that the UVT acrylic is lessstabilized than the MP acrylic The YI rate was also much higher in UVT than MP in the QUV

Fig 7 (a) Average induced absorbance to dose (IAD) of ultraviolet transmitting (UVT) grade poly(methyl methacrylate) (PMMA) after exposure in the QUV accelerated weathering tester to UVA-340 irradiance and after exposure in the Newport solar simulator to AM 15 irradiance for equiva-lent dose steps (b) Average IAD of multipurpose (MP) grade PMMA after exposure in the QUVaccelerated weathering tester to UVA-340 irradiance and after exposure in the Newport solarsimulator to AM 15 irradiance for equivalent dose steps




exposure Region 2 and 2prime could not be compared because there are not significant peaks in UVTrelated to the stabilizer package

54 Yellowing

Changes in yellowness were observed in both acrylic formulations and appear to behave in anear-linear fashion (Fig 6) Changes in yellowness of these materials show the power of YI as aresponse because it is very sensitive to small changes in the optical performance of these materi-als after exposure to these stresses Yellowing was observed in UVTandMP acrylic for these twotypes of exposures Hazing of the samples was not observed Miller et al suggested that yellow-ness and haze are two degradation mechanisms that occur in different classes of PMMAacrylic53

The yellowness that occurred in the samples in the QUV exposure occurred throughout theentire sample after the first dose of exposure The samples in the Newport solar simulator showedthe yellowness occurring predominantly on the first surface closest to the irradiation after the firstdose The yellowing proceeded from the first surface to the interior of the 3 mm sample after eachsubsequent dose This result agrees with the findings that yellowness occurs at the top surface54

and can be reduced with polishing29 The increased depth penetration of yellowing products insamples exposed in the QUV may be due to the fact that samples were backed with standardaluminum sample holders which may have contributed to degradation of the samples Thisallowed some light to be reflected from the aluminum holders through the samples thoughthe extent and wavelength dependence of this contribution has not been thoroughly investigatedAnother possibility for the increase in yellowing in the QUV UVA-340 irradiance samples is thehigher temperature of the stress conditions It has been suggested that there is a relationshipbetween UV radiation and temperature when discussing yellowing and that an increase in tem-perature can increase the yellowing rate16 The samples in the Newport solar simulator wereplaced on a fused ultra-high purity silica photomask blank55 therefore there was very littlereflection of light back through the sample and the temperature of these exposures was lessthan that of the QUV exposures The equivalence of UV exposure of the two different sourceswould also be affected by filter and burner aging in the xenon arc lamp56

55 Stabilizer

While neither of these materials is guaranteed to have a 25-year lifetime in use at 1 sun irradiancefor PVapplications the range of their response to stressors can inform material selection criteriafor PVapplications UVT acrylic lacking UVabsorbers can provide insights about the degrada-tion of the base PMMA resin while MP acrylic shows the effect that a moderate stabilizationpackage can provide The yellowing rates of these two materials show this sharp contrast andMP acrylic shows a 10-fold reduction in yellowing rate Similar observations can be made aboutresponse near the fundamental edge where MP acrylic shows an approximate 10-fold reductionin photodarkening near the fundamental absorption edge

56 Response Ratios and Acceleration

When comparing the ratio of the IADs between the two exposures of equivalent doses the QUVexposure has a much higher acceleration rate of degradation compared to the Newport solarsimulator exposure (Table 4) The acceleration rate in the significant regions for each gradeof PMAA is different which can be caused by test conditions composition of material additivesdegradation mechanism as well as temperature The observation that a constant ldquoaccelerationrdquofactor was not observed in the different degradation regions (1 2 3) and mechanisms demon-strates that the traditional search for a single acceleration factor applicable to a particular materialexposed under differing stresses or stress levels is probably an unreasonable expectation Thissupports the need for developing a more versatile RethSTHORN framework in which to encompassstressors stress levels and multiple responses




57 Stress-Response Framework

Use of a RethSTHORN framework is an important way to study understand and cross-correlate lifetimeand degradation studies of materials components or systems By studying degradationresponses due to multiple types of exposure conditions and steps we are able to understandlinear sub- and supra-linear and nonlinear responses over a wide range of stress levels andtypes of stressors along with the combination of stresses and time-varying stresses Also bycomparing or cross-correlating the effects of multiple stressors to the response of a single stressexposure we can understand the specific (combined or sequential) synergistic mechanisms atplay in a material system Significant understanding of the ways a material may behave over itsservice life can be understood by comparing results of real-world exposures to the stress andresponse framework of the system that has been developed beforehand In cases where unwanteddegradation is observed the stress and response framework can be used to understand and guidemitigation of these unwanted responses Additionally because a stress and response frameworkcan include highly accelerated stress conditions without assuming that these conditions maplinearly to in-service exposures useful information about end-of-life performance can begleaned505758

6 Conclusion

Two grades of acrylic PMMA with different concentration of UV stabilizers were exposed tostresses in order to understand the stress-response framework for PMMA acrylic By using IADand YI as response metrics and comparing these responses across material systems in the sameclass and across different stress conditions valuable insights into the lifetime performance ofthese materials can be understood Both MP and UVT acrylic show similarities in their degrada-tion with exposure to simulated sunlight and UVA-340 fluorescent weathering The materialsrespond to the fluorescent exposure at a higher rate than in the simulated solar exposure how-ever this variation may be due to differences in sample temperature during exposure The QUVexposure showed to be a more aggressive weathering test than exposure to the Newport solarsimulator being three to six times more damaging to UVT acrylic and one to three times moredamaging to MP acrylic on a per-dose basis QUV showed more yellowing while a ten-foldreduction of these phenomena was observed with the addition of the stabilizer package inMP acrylic Exposure doses achieved were equivalent to one to eight years of dose when com-pared to the UV-only and full-spectrum dose of AM 15D in Arizona while these tests wereperformed in less than 40 days Further study into degradation mechanisms of acrylic will elu-cidate the exact phenomena that contribute to the material responses to stress This frameworkand these insights are being applied to other acrylic grades for Fresnel lenses and mirrors

Continuing work will require identification and quantization of stressors identification ofways to meaningfully accelerate stresses and reproduce naturally occurring degradation mechan-isms and application of this framework to other stressors such as temperature humidity mechan-ical load corrosive environments and applied loads An understanding of the stress-responseframework in the PV LampDS will allow for the same framework to be applied to real-worldenvironmental conditions to enable indoor testing to be correlated to outdoor lifetimes

Further work will entail the monitoring of other responses such as haze bidirectional scat-tering distribution function and contact angle measurements of these two PMMA grades as well

Table 4 Ratios of average IAD values from QUV (step 4) to Newport solar simulator (step 1) foreach region for the two grades of PMMA of equivalent dose steps

Region 1 (275 nm) Region 2 (298 nm) Region 2prime (339 nm) Region 3 (440 nm)

UVT 28 33 46 63

MP 12 16 086 38




as other grades of PMMA after stress exposure In addition dark recovery time and humiditywill be investigated in a multifactor stress experiment

Acknowledgments

The authors acknowledge funding from Ohio Third Frontier Photovoltaic Program Award Tech11-060 Research was performed at the SDLE Center at Case Western Reserve Universityfunded through the Ohio Third Frontier Wright Project Program Award Tech 12-004 Weappreciate the diligent efforts of Esther Deena at the SDLE center Mark Schuetz of ReplexPlastics Sean Fowler of Q-Labs Sample tracking and data management support was providedby REDCapTM59 funded by the Clinical and Translational Science Collaborative (CTSC) grantsupport (1 UL1 RR024989 from NCRRNIH)

References

1 US DOE ldquoWorkshop on Science for Energy Technology workshop report for DOEBasic Energy Science Advisory Committeerdquo (August 2010) httpscienceenergygov~mediabespdfreportsfilessetf_rptpdf

2 M P Murray et al ldquoSolar radiation durability framework applied to acrylic solar mirrorsrdquoProc SPIE 8112 811203 (2011) httpdxdoiorg10111712893827

3 R H French et al ldquoSolar radiation durability of materials components and systems for lowconcentration photovoltaic systemsrdquo in 2011 IEEE Energytech pp 1ndash5 IEEE ClevelandOH (2011)

4 R H French et al ldquoMaterials for concentrator photovoltaic systems optical properties andsolar radiation durabilityrdquo in 6th International Conference on Concentrating PhotovoltaicSystems (cpv-6) 1277 A W Betts F Dimroth R D McConnell and G Sala Edspp 127ndash130 Amer Inst Physics Melville (2010)

5 R H French et al ldquoOptical properties of polymeric materials for concentrator photovoltaicsystemsrdquo Sol Energy Mater Sol Cells 95(8) 2077ndash2086 (2011) httpdxdoiorg101016jsolmat201102025

6 M P Murray and R H French ldquoSolar radiation durability of materials components andsystems for photovoltaicsrdquo in 2011 37th IEEE Photovoltaic Specialists Conference (PVSC)pp 000972ndash000977 IEEE Seattle WA (2011) M P Murray et al ldquoDegradation of backsurface acrylic mirrors for low concentration and mirror-augmented photovoltaicsrdquo ProcSPIE 8472 847205 (2012) httpdxdoiorg10111712930102

7 M P Murray L S Bruckman and R H French ldquoDurability of acrylic stress and responsecharacterization of materials for photovoltaicsrdquo in 2012 IEEE Energytech pp 1ndash6 IEEECleveland OH (2012)

8 M White and J B Bernstein ldquoMicroelectronics reliability physics-of-failure based mod-eling and lifetime evaluationrdquo p 210 JPL Publication 08-5 07-102 Pasadena CA (2008)

9 UL 1703 ldquoUL Standard for Safety Flat-Plate Photovoltaic Modules and PanelsrdquoUnderwriters Laboratories (2011)

10 IEC 62108 ldquoConcentrator photovoltaic (CPV) modules and assembliesmdashdesign qualifica-tion and type approvalrdquo Int Electrotechnical Comm (2007)

11 IEC 61215 ldquoCrystalline silicon terrestrial photovoltaic (PV) modulesmdashdesign qualificationand type approvalrdquo Int Electrotechnical Comm (2005)

12 IEC 61646 ldquoThin-film terrestrial photovoltaic (PV) modulesmdashdesign qualification andtype approvalrdquo Int Electrotechnical Comm (2008)

13 J W McPherson Reliability Physics and Engineering Time-to-Failure Modeling SpringerScience+Business Media LLC New York NY (2010)

14 M Pecht Product Reliability Maintainability and Supportability Handbook 2nd edCRC Press Boca Raton FL (2009)

15 K Branker M J M Pathak and J M Pearce ldquoA review of solar photovoltaic levelized costof electricityrdquo Renew Sustain Energ Rev 15(9) 4470ndash4482 (2011) httpdxdoiorg101016jrser201107104




16 D C Miller et al ldquoDurability of poly(methyl methacrylate) lenses used in concentratingphotovoltaic modulesrdquo Proc SPIE 7773 777303 (2010) httpdxdoiorg10111712861096

17 M Poliskie Solar Module Packaging Polymeric Requirements and Selection CRC PressBoca Raton FL (2011)

18 BASF Corporation 205 South James Street Newport DE19 J E Pickett and J E Moore ldquoPhotodegradation of UV screeners polymer degradation and

stabilityrdquo Polymer Degrad Stabil 42(3) 231ndash244 (1993) httpdxdoiorg1010160141-3910(93)90219-9

20 BLSreg 5411 Ultraviolet Light Absorber amp Stabilizer httpwwwmayzocompdfBLS5411pdf

21 A Charlesbey and D K Thomas ldquoA comparison of the effects of ultra-violet and gammaradiation in polymethylmethacrylaterdquo Proc Roy Soc A 269(1336) 104ndash124 (1962) httpdxdoiorg101098rspa19620165

22 M Abouelezz and P F Waters ldquoStudies on the photodegradation of poly(methyl metha-crylate)rdquo NBSIR 78-1463 National Bureau of Standards pp 1ndash55 (1978)

23 A R Shultz ldquoDegradation of polymethyl methacrylate by ultraviolet lightrdquo J Phys Chem65(6) 967ndash972 (1961) httpdxdoiorg101021j100824a019

24 R B Fox L G Isaacs and S Stokes ldquoPhotolytic degradation of poly(methyl methacrylate)rdquoJ Polym Sci A 1(3) 1079ndash1086 (1963) httpdxdoiorg101002pol1963100010321

25 R B Fox et al ldquoPhotodegradation of poly(methyl methacrylate)rdquo J Polym Sci A 2(5)2085ndash2092 (1964) httpdxdoiorg101002pol1964100020506

26 A R Mahoney J E Cannon and J R Woodworth ldquoAccelerated UV-aging of acrylicmaterials used in PV concentrator systemsrdquo in Conference Record of the Twenty ThirdIEEE Photovoltaic Specialists Conference pp 1216ndash1221 IEEE Louisville KY (1993)

27 A Torikai M Ohno and K Fueki ldquoPhotodegradation of poly(methyl methacrylate) bymonochromatic light quantum yield effect of wavelengths and light intensityrdquo J ApplPolym Sci 41(5ndash6) 1023ndash1032 (1990) httpdxdoiorg101002app1990070410513

28 T Mitsuoka A Torikai and K Fueki ldquoWavelength sensitivity of the photodegradation ofpoly(methyl methacrylate)rdquo J Appl Polym Sci 47(6) 1027ndash1032 (1993) httpdxdoiorg101002app1993070470609

29 A Torikai S Hiraga and K Fueki ldquoPhotodegradation of blends of poly(methyl metha-crylate) and poly(styrene-co-methyl methacrylate)rdquo Polym Deg Stab 37(1) 73ndash76 (1992)httpdxdoiorg1010160141-3910(92)90094-L

30 L G Rainhart and W P Schimmel ldquoEffect of outdoor aging on acrylic sheetrdquo Solar Energ17(4) 259ndash264 (1975) httpdxdoiorg1010160038-092X(75)90008-0

31 J W Summers and E B Rabinovitch ldquoWeather ability of vinyl and other plasticsrdquo inWeathering of Plastics Testing to Mirror Real Life Performance G Wypch Ed WilliamAndrew Publishing Norwich (1999)

32 A Torikai T Hattori and T Eguchi ldquoWavelength effect on the photoinduced reaction ofpolymethylmethacrylaterdquo J Polym Sci A 33(11) 1867ndash1871 (1995) httpdxdoiorg101002pola1995080331114


34 T Caykara and O Guumlven ldquoUV degradation of poly(methyl methacrylate) and its vinyl-triethoxysilane containing copolymersrdquo Polym Deg Stab 65(2) 225ndash229 (1999)httpdxdoiorg101016S0141-3910(99)00008-7

35 M Shirai T Yamamoto and M Tsunooka ldquoAblative photodegradation of poly(methylmethacrylate) and its homologues by 185 nm lightrdquo Polym Deg Stab 63(3) 481ndash487(1999) httpdxdoiorg101016S0141-3910(98)00077-9

36 I Mita K Obata and K Horie ldquoPhotoinitiated thermal degradation of polymers II poly(methyl methacrylate)rdquo Polym J 22(5) 397ndash410 (1990) httpdxdoiorg101295polymj22397

37 M Kato and Y Yoneshige ldquoPhotodegradation behavior of some vinyl ketone copolymersrdquoDie Makromolekulare Chemie 164(1) 159ndash169 (1973) httpdxdoiorg101002macp1973021640116




38 R B Fox and L G IsaacsPhotodegradation of High Polymers Part IiimdashPhotolysis ofPoly(Methyl methacrylate) in Vacuum and in Air in NLR-5720 Naval Research LabWashington DC (1961)

39 ASTM G173mdash03 Standard Tables for Reference Solar Spectral Irradiances Direct Nor-mal and Hemispherical on 37deg Tilted Surface ASTM International West Conshohocken(2008)

40 Q-Lab Q-Lab Headquarters amp Instruments Division Cleveland OH41 ASTM G154 ldquoStandard Practice for Operating Fluorescent Light Apparatus for UV

Exposure of Nonmetallic Materialsrdquo ASTM International West Conshohocken PA(2005)

42 Replex Plastics Inc Mount Vernon OH wwwreplexcom43 Daylighting (March 2012) httpwwwwikipediacomdaylighting44 National Solar Radiation Data Base ldquo1991ndash2005 update typical meteorological yearrdquo

httprredcnrelgovsolarold_datansrdb1991-2005tmy345 D Kockett ldquoFactors influencing the reliability of results in accelerated weathering testsrdquo

in Durability Testing of Nonmetallic Materials ASTM STP 1294 R J Herling EdAmerican Society for Testing and Materials (1996)

46 Newport Corporation North Billerica MA httpwwwnewportcom47 FLIR Systems Americarsquos Main Office Boston MA httpwwwflircom48 H V Tran et al ldquoFluidmdashphotoresist interactions and imaging in high index immersion

lithographyrdquo J Microlith Microfab Microsyst 8(3) 033006 (2009) httpdxdoiorg10111713224950

49 R H French and H V Tran ldquoImmersion lithography photomask and wafer-level materi-alsrdquo Ann Rev Mater Res 39 93ndash126 (2009) httpdxdoiorg101146annurev-matsci-082908-145350

50 ASTM E313 Standard Practice for Calculating Yellowness and Whiteness Indices fromInstrumentally Measured Color Coordinates ASTM International West ConshohockenPA (2005)

51 J E Pickett ldquoHighly predictive accelerated weathering of engineering thermoplasticsrdquoAtlas Mater Test Product Technol News 35(73) 1ndash10 (2005)

52 W J Putnam M McGreer and D Pekara ldquoParametric control of Fresnel reflecting concen-trator outdoor accelerated weathering devicerdquo Durability Testing of Nonmetallic MaterialsASTM STP 1294 R J Herling Ed American Society for Testing and Materials (1996)

53 D C Miller et al ldquoThe durability of poly(methyl methacrylate) lenses used in concentrat-ing photovoltaic technologyrdquo in ATLASNIST Workshop on Photovoltaic Materials Relia-bility (2011)

54 D C Miller and S R Kurtz ldquoDurability of Fresnel lenses a review specific to the con-centrating photovoltaic applicationrdquo Sol Energy Mater Sol Cells 95(8) 2037ndash2068(2011) httpdxdoiorg101016jsolmat201101031

55 G L Tan et al ldquoOptical properties and London dispersion interaction of amorphous andcrystalline SiO2 determined by vacuum ultraviolet spectroscopy and spectroscopic ellipso-metryrdquo Phys Rev B 72(20) 205117 (2005) httpdxdoiorg101103PhysRevB72205117

56 W D Ketola T S Skogland and R M Fischer ldquoEffects of filter and burner aging on thespectral power distribution of xenon arc lampsrdquo inDurability Testing on Nonmetallic Mate-rials ASTM STP 1294 R J Herling Ed American Society for Testing and Materials(1996)

57 H V Tran et al ldquoHigh refractive index fluid evaluations at 193 nm fluid lifetime and fluidresist interaction studiesrdquo J Photopolym Sci Technol 21(5) 631ndash639 (2008) httpdxdoiorg102494photopolymer21631

58 R H French et al ldquoHigh index immersion lithography with second generation immersionfluids to enable numerical apertures of 155 for cost effective 32 nm half pitchesrdquo ProcSPIE 6520 652010 (2007) httpdxdoiorg10111712712234

59 P A Harris et al ldquoResearch electronic data capture (REDCap)mdasha metadata driven meth-odology and workflow process for providing translational research informatics supportrdquoJ Biomed Inform 42(2) 377ndash381 (2009) httpdxdoiorg101016jjbi200808010




Myles P Murray is a graduate student at Case Western Reserve Universityin the Department of Materials Science and Engineering He received aBachelor of Arts in chemistry from Hendrix Collegersquos School for GreenEnvironmental Chemistry Before attending CWRU he was an environ-mental monitoring specialist with the Division of Air Quality in ClevelandOhio His research is focused on improving the cost of energy from of solarenergy through enhanced optical materials for photovoltaic systems andimproving bankability of PV systems through lifetime and degradationscience

Laura S Bruckman is a research associate at CWRU She received herPhD in analytical chemistry from the University of South Carolina in DrMichael L Myrickrsquos research group in August 2011 Her research while atUniversity of South Carolina was on the rapid classification of phytoplank-ton cells using single-cell fluorescence measurements and multivariateoptical computing Her work since joining Dr Roger Frenchrsquos group atCWRU has been on the degradation mechanisms of acrylic and acrylicback surface mirrors in order to better understand the Lifetime andDegradation Science of these materials

Roger H French (F Alex Nason Professor Department of MaterialsScience and Engineering) joined CWRU in August 2010 after 24 yearsof conducting basic research and product development in DuPontrsquos CentralResearch He is the director of the Solar Durability and Lifetime ExtensionCenter at CWRU He has a broad experience in developing and commer-cializing optical materials for many different applications and in optimizingthese materials for improved radiation durability and lifetime He is also anationally recognized expert in Lifetime and Degradation Science (LampDS)for commercial applications evidenced by his work on attenuating phase

shift photomasks fluoropolymer pellicles for photolithography immersion lithography imagingfluids and materials for concentrating photovoltaic systems He has 22 issued patents and morethan 145 publications




Photodegradation in a stress and response frameworkpoly(methyl methacrylate) for solar mirrors and lens

Myles P Murray Laura S Bruckman and Roger H FrenchCase Western Reserve University Solar Durability and Lifetime Extension Center and

Materials Science Department Cleveland Ohio 44106rogerfrenchcaseedu

Abstract In the development of materials for enhanced photovoltaic (PV) performance it iscritical to have quantitative knowledge of both their initial performance and their performanceover the required 25-year warranted lifetime of the PV system Lifetime and degradation sciencebased on an environmental stress and response framework is being developed to link the inten-sity and net stress to which materials components and systems are exposed to the responsesobserved and their subsequent degradation and damage accumulation over the lifetime Inducedabsorbance to dose (IAD) a metric developed for solar radiation durability studies of solar andenvironmentally exposed materials is defined as the rate of photodarkening or photobleaching ofa material as a function of radiation dose Quantitative degradation rates like IAD determinedover a wide range of stress intensities and net stresses have the potential to predict degradationfailure and power loss rates in photovoltaic systems over time caused by damage accumulationTwo grades of poly(methyl methacrylate) were exposed and evaluated in two cases of high-intensity ultraviolet exposures A three- to six-fold increase in photodarkening was observedfor one acrylic formulation when exposed to UVA-340 light when compared with concentratedxenon-arc exposure The other more highly stabilized acrylic formulation showed up to threetimes more photodarkening in the same exposure copy 2012 Society of Photo-Optical InstrumentationEngineers (SPIE) [DOI 1011171JPE2022004]

Keywords photodegradation acrylic poly(methyl methacrylate) photovoltaics degradation

Paper 12020SS received Mar 21 2012 revised manuscript received Jul 26 2012 accepted forpublication Oct 1 2012 published online Nov 19 2012

1 Introduction

A recent US Department of Energyworkshop on Science for EnergyTechnologies1 identified thetopic of photovoltaics (PV) lifetime and degradation science (LampDS)2ndash4 as a critical scientificchallenge for robust adoption of PV The 25-year lifetime performance of PV requires a betterunderstanding of the degradation mechanisms in PV materials components and systems Bydeveloping metrics metrology and tools to quantify compare and cross-correlate the responseof PV systems components and materials to a variety of stressors such as ultraviolet (UV) radia-tion humidity and temperature variation for both accelerated and outdoor testing we can linkstresses to observed responses in a stress-response framework and determine quantitative rates ofdegradation LampDS requires the development of quantitative degradation mechanisms and ratesfor degradation and failure modes so as to enable quantitative lifetime projections

Lifetime and degradation science (LampDS) based on a stress and response [RethSTHORN] frameworkis being developed that links the intensity and net stress to which materials components andsystems are exposed to the responses observed and the degradation and damage accumulationover lifetime This RethSTHORN framework can encompass multifactor and cyclic environmental stres-sors including solar irradiance temperature and humidity which can cause degradation overtime The RethSTHORN framework allows for the cross-correlation of the materials response and degra-dation rates with numerous applied stress intensities and net stresses and allows determination ofthe effects of stressors even at accelerated intensities with applicability to multiple service con-ditions At the same time an RethSTHORN framework permits incorporation of data determined over

0091-32862012$2500 copy 2012 SPIE
























S frac14Z

σdt (1)

XS frac14





fethσTHORNσdt (3)





3 Methods












Ee frac14Z

360

280

Eeλdλ (5)





UVA-340 280 to 360 nm 576 W∕m2 6065 W∕m2

TUV 280 to 400 nm 1570 W∕m2 845 W∕m2



UVA-340 280 to 360 nm 151 GJ∕m2 206 MJ∕m2

TUV 280 to 400 nm 244 GJ∕m2 287 MJ∕m2

Time (days) 18 39





He frac14Z

Eedt frac14Z

t

0

Z360

280

Eeλdλ (6)




33 Evaluations





333 Temperature


34 Metrics







AverageAbs

cmper

GJ


Doset(7)

IncrementalAbs

cmper

GJ








(9)




4 Results







41 IAD Results





ExposuresPMMAgrade

Region 1(275 nm)

Region 2(298 nm)


Region 3(440 nm)

QUV UVT 0019 00069 00029 000070



MP 00036 000045 minus000033 18 e-5






43 Temperature







5 Discussion








QUV 45 99 117

NSS 19 61 51















54 Yellowing




55 Stabilizer









6 Conclusion






UVT 28 33 46 63

MP 12 16 086 38





Acknowledgments


References































































































S frac14Z

σdt (1)

XS frac14





fethσTHORNσdt (3)





3 Methods












Ee frac14Z

360

280

Eeλdλ (5)





UVA-340 280 to 360 nm 576 W∕m2 6065 W∕m2

TUV 280 to 400 nm 1570 W∕m2 845 W∕m2



UVA-340 280 to 360 nm 151 GJ∕m2 206 MJ∕m2

TUV 280 to 400 nm 244 GJ∕m2 287 MJ∕m2

Time (days) 18 39





He frac14Z

Eedt frac14Z

t

0

Z360

280

Eeλdλ (6)




33 Evaluations





333 Temperature


34 Metrics







AverageAbs

cmper

GJ


Doset(7)

IncrementalAbs

cmper

GJ








(9)




4 Results







41 IAD Results





ExposuresPMMAgrade

Region 1(275 nm)

Region 2(298 nm)


Region 3(440 nm)

QUV UVT 0019 00069 00029 000070



MP 00036 000045 minus000033 18 e-5






43 Temperature







5 Discussion








QUV 45 99 117

NSS 19 61 51















54 Yellowing




55 Stabilizer









6 Conclusion






UVT 28 33 46 63

MP 12 16 086 38





Acknowledgments


References























































































S frac14Z

σdt (1)

XS frac14





fethσTHORNσdt (3)





3 Methods












Ee frac14Z

360

280

Eeλdλ (5)





UVA-340 280 to 360 nm 576 W∕m2 6065 W∕m2

TUV 280 to 400 nm 1570 W∕m2 845 W∕m2



UVA-340 280 to 360 nm 151 GJ∕m2 206 MJ∕m2

TUV 280 to 400 nm 244 GJ∕m2 287 MJ∕m2

Time (days) 18 39





He frac14Z

Eedt frac14Z

t

0

Z360

280

Eeλdλ (6)




33 Evaluations





333 Temperature


34 Metrics







AverageAbs

cmper

GJ


Doset(7)

IncrementalAbs

cmper

GJ








(9)




4 Results







41 IAD Results





ExposuresPMMAgrade

Region 1(275 nm)

Region 2(298 nm)


Region 3(440 nm)

QUV UVT 0019 00069 00029 000070



MP 00036 000045 minus000033 18 e-5






43 Temperature







5 Discussion








QUV 45 99 117

NSS 19 61 51















54 Yellowing




55 Stabilizer









6 Conclusion






UVT 28 33 46 63

MP 12 16 086 38





Acknowledgments


References










































































S frac14Z

σdt (1)

XS frac14





fethσTHORNσdt (3)





3 Methods












Ee frac14Z

360

280

Eeλdλ (5)





UVA-340 280 to 360 nm 576 W∕m2 6065 W∕m2

TUV 280 to 400 nm 1570 W∕m2 845 W∕m2



UVA-340 280 to 360 nm 151 GJ∕m2 206 MJ∕m2

TUV 280 to 400 nm 244 GJ∕m2 287 MJ∕m2

Time (days) 18 39





He frac14Z

Eedt frac14Z

t

0

Z360

280

Eeλdλ (6)




33 Evaluations





333 Temperature


34 Metrics







AverageAbs

cmper

GJ


Doset(7)

IncrementalAbs

cmper

GJ








(9)




4 Results







41 IAD Results





ExposuresPMMAgrade

Region 1(275 nm)

Region 2(298 nm)


Region 3(440 nm)

QUV UVT 0019 00069 00029 000070



MP 00036 000045 minus000033 18 e-5






43 Temperature







5 Discussion








QUV 45 99 117

NSS 19 61 51















54 Yellowing




55 Stabilizer









6 Conclusion






UVT 28 33 46 63

MP 12 16 086 38





Acknowledgments


References














































































Ee frac14Z

360

280

Eeλdλ (5)





UVA-340 280 to 360 nm 576 W∕m2 6065 W∕m2

TUV 280 to 400 nm 1570 W∕m2 845 W∕m2



UVA-340 280 to 360 nm 151 GJ∕m2 206 MJ∕m2

TUV 280 to 400 nm 244 GJ∕m2 287 MJ∕m2

Time (days) 18 39





He frac14Z

Eedt frac14Z

t

0

Z360

280

Eeλdλ (6)




33 Evaluations





333 Temperature


34 Metrics







AverageAbs

cmper

GJ


Doset(7)

IncrementalAbs

cmper

GJ








(9)




4 Results







41 IAD Results





ExposuresPMMAgrade

Region 1(275 nm)

Region 2(298 nm)


Region 3(440 nm)

QUV UVT 0019 00069 00029 000070



MP 00036 000045 minus000033 18 e-5






43 Temperature







5 Discussion








QUV 45 99 117

NSS 19 61 51















54 Yellowing




55 Stabilizer









6 Conclusion






UVT 28 33 46 63

MP 12 16 086 38





Acknowledgments


References










































































He frac14Z

Eedt frac14Z

t

0

Z360

280

Eeλdλ (6)




33 Evaluations





333 Temperature


34 Metrics







AverageAbs

cmper

GJ


Doset(7)

IncrementalAbs

cmper

GJ








(9)




4 Results







41 IAD Results





ExposuresPMMAgrade

Region 1(275 nm)

Region 2(298 nm)


Region 3(440 nm)

QUV UVT 0019 00069 00029 000070



MP 00036 000045 minus000033 18 e-5






43 Temperature







5 Discussion








QUV 45 99 117

NSS 19 61 51















54 Yellowing




55 Stabilizer









6 Conclusion






UVT 28 33 46 63

MP 12 16 086 38





Acknowledgments


References











































































AverageAbs

cmper

GJ


Doset(7)

IncrementalAbs

cmper

GJ








(9)




4 Results







41 IAD Results





ExposuresPMMAgrade

Region 1(275 nm)

Region 2(298 nm)


Region 3(440 nm)

QUV UVT 0019 00069 00029 000070



MP 00036 000045 minus000033 18 e-5






43 Temperature







5 Discussion








QUV 45 99 117

NSS 19 61 51















54 Yellowing




55 Stabilizer









6 Conclusion






UVT 28 33 46 63

MP 12 16 086 38





Acknowledgments


References











































































41 IAD Results





ExposuresPMMAgrade

Region 1(275 nm)

Region 2(298 nm)


Region 3(440 nm)

QUV UVT 0019 00069 00029 000070



MP 00036 000045 minus000033 18 e-5






43 Temperature







5 Discussion








QUV 45 99 117

NSS 19 61 51















54 Yellowing




55 Stabilizer









6 Conclusion






UVT 28 33 46 63

MP 12 16 086 38





Acknowledgments


References












































































43 Temperature







5 Discussion








QUV 45 99 117

NSS 19 61 51















54 Yellowing




55 Stabilizer









6 Conclusion






UVT 28 33 46 63

MP 12 16 086 38





Acknowledgments


References










































































5 Discussion








QUV 45 99 117

NSS 19 61 51















54 Yellowing




55 Stabilizer









6 Conclusion






UVT 28 33 46 63

MP 12 16 086 38





Acknowledgments


References





















































































54 Yellowing




55 Stabilizer









6 Conclusion






UVT 28 33 46 63

MP 12 16 086 38





Acknowledgments


References











































































54 Yellowing




55 Stabilizer









6 Conclusion






UVT 28 33 46 63

MP 12 16 086 38





Acknowledgments


References












































































6 Conclusion






UVT 28 33 46 63

MP 12 16 086 38





Acknowledgments


References











































































Acknowledgments


References






































































































































































Date post:	22-Jul-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

Photodegradation in a stress and response framework: poly ......topic of photovoltaics (PV) lifetime...

Documents