Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016
energy.gov/sunshotenergy.gov/sunshotenergy.gov/sunshot
CSPProgramSummit2016
Predic'vephysico-chemicalmodelingofintrinsicdegrada'onmechanismsforadvancedreflectormaterials
RossLarsenSeniorScien<stNa<onalRenewableEnergyLaboratory
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 2
Outline
• Projectobjec<ves
• Approach
• Kine<csandtransport
• Mechanicalproper<es
• Agingofreflectorstacks
• Summaryandoutlook
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016
ProjectTeamandAcknowledgments
NREL:RossLarsenRayGroutChaoZhangAndrewFergusonRobertTirawatMaPhewGray(nowwithSkyFuel)VinceWheeler(nowatAustralianNa<onalUniversity)
Costsharepartners:DLR
PeterHellerFlorianSuPer
DOECSP
DE-FOA-0000861
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016
ProjectObjec'ves
• Developnovelmodelingcapabilityforunderstandingreflectordegrada<onthatcombinesareac<on-diffusion-op<csformalismwithexperimentstoiden<fyunderlyingratesandmechanismsfordegrada<on
• Createanddisseminateatool,validatedbytes<ngagainstmul<pleexperiments,thatcanbeusedbytheCSPcommunitytoevaluatedegrada<onmechanisms
Time
Refle
ctance StackdesignsAandB
~30yr
polymer/glasssubstrate
reflec<nglayer
ARcoa<ng
protec<vecoa<ng
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016
Micro-kinetic Description
Intrinsic material property characterization
Transport Optical Damage
Mechanical Property
Accelerated Aging
IndividualComponentLayer
Characterizesinglelayers
orasimplifiedstack
σ c = σ c (nC,nbl ,nOx,nH2O
,T )
∂nC(r,t)∂t
= DC∇2nC(r,t)− α(λ)Iλ (r,t)dλ nC(r,t)− n*(r,t)− nbl (r,t)⎡
⎣⎤⎦∫
∂n*(r,t)∂t
= D*∇2n*(r,t)+ α(λ)Iλ (r,t)dλ∫ nC(r,t)− n*(r,t)⎡
⎣⎤⎦ − kOxnOx(r,t) ⋅n*(r,t)
∂nbl (r,t)∂t
= +kOxnOx(r,t) ⋅n*(r,t)
∂nOx(r,t)∂t
= DOx∇2nOx(r,t)− kOxnOx(r,t) ⋅n*(r,t)+S(r,t)
Iλ (r,t) = Iλ (z = 0,t)exp[−α (λ) nC (z,t)−n* (z,t)⎡⎣ ⎤⎦dz
0
L
∫]
Simplifiedexamplephoto-kine<cscheme
αλ = αλ (nC,nbl ,nOx,nH2O
,T )
D = D(nC,nbl ,nOx,nH2O
,T )
kOx = kOx(nC,nbl ,T )
k = k(T,P)
nOx(0,t) = kPOx
ProjectApproach
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016
ProjectApproach
• Createandvalidatemodelsandsodwaretopredictandunderstanddegrada<onofadvancedreflectorsspanningdecades• Op<cs,reac<on-diffusionkine<cssolver(feedback)• Finiteelementmechanicalmodeling
Op<cs(fullHelmholtz)
Kine<cs/transport(splitoperator)
Mechanicalmodelingdelamina<on,cracking,etc.
(FEM)
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016
Op'cssolver
• Transfermatrixmethodfor1Dreflectorstacks• Fullyvectorial(sandppolariza<on)• Numericallyexactforpiecewiseconstantstacks• Exactreflectancewithoutneedforupwinding,boundaryelements,etc.(3Dversion:boundaryelements)
0
0.2
0.4
0.6
0.8
1
500 1000 1500 2000 2500
Refle
ctan
ce
Wavelength (nm)
500 A Cu/1000 A Ag/1000 A Al_2O_3500 A Cu/2000 A Ag/1000 A Al_2O_3500 A Cu/1000 A Ag/6000 A Al_2O_3500 A Cu/2000 A Ag/6000 A Al_2O_3
Unpolarizedreflectancefromalumina-coatedsilverreflector
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016
Kine'cs-transportsolver
TransientmeasurementofpermeabilityofPETfilmtowater
Delayedonsetconsistentwithnon-FickiandiffusionDualsorp<on:
!"!" = !!
!!"
11+ ! !!!!
(1+ !!!!)!
!"!"
Non-linearityintransportandsorp<onrequiresnovelsolverforboundaries
Figure 3: Four domains sharing a singe interface in the structured finite vol-ume formulation. Many domains meeting at a single interface adds little or nodi�culty to the use of the proposed interface residual method.
To handle an arbitrary number of subdomains, we first recall that we canwrite a set of equations of the form (50) for the ith subdomain,
"A
(i)II
A
(i)Ig
0 I
# u
(i)I
u
(i)g
!=
f
(i)I
u
(i)ng
!. (50)
So we have the interior nodes in terms of the ghost nodes,
u
(i)I
= �A
(i)�1II
A
(i)Ig
u
g
+ A
(i)�1II
f
(i)I
. (93)
The interface residual can be set up on a pairwise basis. From Figure 3 we
18
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016
Kine'cs-transportsolver
TransientmeasurementofpermeabilityofPETfilmtowater
Delayedonsetconsistentwithnon-FickiandiffusionDualsorp<on:
!"!" = !!
!!"
11+ ! !!!!
(1+ !!!!)!
!"!"
Wehaveextractedpermeabilityasafunc<onofRH,TandPox,TforPET,PMMA,withandwithoutoxidecoa<ngs
P= !! + !!! !!!/!
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 10CSPProgramSummit2016
0
20
40
60
80
100
120
140
0 5 10 15 20 25 30
Stre
ss (M
Pa)
Strain (%)
0h UV Exposure, Direction 1, Sample 1 0h UV Exposure, Direction 1, Sample 2 0h UV Exposure, Direction 2, Sample 1 0h UV Exposure, Direction 2, Sample 2 68h UV Exposure, Direction 1, Sample 1 68h UV Exposure, Direction 1, Sample 2 68h UV Exposure, Direction 2, Sample 1 68h UV Exposure, Direction 2, Sample 2 78h UV Exposure, Direction 1, Sample 1 78h UV Exposure, Direction 1, Sample 2 78h UV Exposure, Direction 2, Sample 1 78h UV Exposure, Direction 2, Sample 2
Mechanicalproper'es:UVdegrada'onofpoly(ethyleneterephthalate)films
2.0
1.5
1.0
0.5
0.0
Abso
rban
ce
500450400350300250
Wavelength (nm)
0 hrs (un-aged) 0.5 hrs 1 hr 2 hrs 3 hrs 4 hrs 5 hrs 6 hrs 7 hrs 8 hrs 9 hrs 10 hrs 12 hrs 14 hrs 16 hrs 20 hrs 32 hrs 37.5 hrs 48 hrs 55.25 hrs 80 hrs 96 hrs 125.5 hrs 160 hrs 191.25 hrs 229.25 hrs 302.25 hrs
10XUVA/UVBunderambientcondi<ons
Agedfilmsshowminimalchangeinelas<cconstants,drama<creduc<onin
ul<matefailurestrain
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 11CSPProgramSummit2016
Mechanicalproper'es:FractureofbriIlecoa'ngsonstrainedPETfilms
11
• Tensile properties of TiO2/PET thin film system
0
20
40
60
80
100
120
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
Coating thickness of 100 nm
Coating thickness of 300 nm
Coating thickness of 450 nm
Coating thickness of 600 nm
Pure substrate
Stre
ss (M
Pa)
Strain
y = 0.0083x + 3728
3550
3600
3650
3700
3750
3800
3850
3900
3950
0 100 200 300 400 500 600 700
Elastic modulus
Linear data fitting
Elas
tic M
odul
us (M
Pa)
Coating thickness (nm)
(a) (b)
Ø Tensile load is transferred through the polymer substrate.
Ø Coated sample shows an earlier yielding.
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 12CSPProgramSummit2016
Mechanicalproper'es:FractureofbriIlecoa'ngsonstrainedPETfilms
12
• In situ tensile test shows progressive cracking behavior
50 µm
2.5%
9.9%
12.8%
18.2%
Tensile direction
Parallel crack
Buckle
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 13CSPProgramSummit2016
Mechanicalproper'es:FractureofbriIlecoa'ngsonstrainedPETfilms
13
• Tape test shows delamination of TiO2 for UV aged sample
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 14CSPProgramSummit2016
Mechanicalproper'es:FractureofbriIlecoa'ngsonstrainedPETfilms
14
• In situ tensile test shows progressive cracking behavior
50 µm
2.5%
9.9%
12.8%
18.2%
Tensile direction
Parallel crack
Buckle Howdofractureanddelamina<onchangewithaging?
Isthereanunderlyingkine<cmodel?
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 15CSPProgramSummit2016
Agingofsimplifiedreflectorstacks
15
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 16CSPProgramSummit2016
Agingofsimplifiedreflectorstacks:Un-aged
16
100
80
60
40
20
0
Refle
ctan
ce (
%)
15001000500wavelength (nm)
Silver (Reflectance)Literature
Adachi Johnson Schulz
Measured
Front-surface Al standard Predicted from VASE-derived Optical Constants
Stored in Air Spot 1 Spot 2 Spot 3
Cau<on:thereisnotnecessarily“canonical”silver
Ourmodelsnowhave“thin”and“thick”NRELsilverop<calconstants
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 17CSPProgramSummit2016
Agingofsimplifiedreflectorstacks:Un-coated
17
Degradesamplesundercondi<onsof45C,40%RH2Xsunlight(ordark)HemisphericalR
0
10
20
30
40
50
60
70
80
90
100
250 750 1250 1750 2250
Refle
ctan
ce
Wavelength(nm)
5592-28-2,Cu(10)/Ag(500)
5592-28-2-1
5592-28-2-130hours
5592-28-2-160hours
5592-28-2-2
5592-28-2-230hours
5592-28-2-260hours
5592-28-2-290hours
5592-28-2-3
5592-28-2-330hours
5592-28-2-360hours
5592-28-2-390hours
5592-28-2-4
5592-28-2-430hours
5592-28-2-460hours
5592-28-2-490hours
Substrates/Condi<on1-Glass/Light2-Glass/Dark3-3milPET/Light4-3milPET/Dark
Silverdegradesdifferentlyinthedarkthanunderillumina<on
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 18CSPProgramSummit2016
Agingofsimplifiedreflectorstacks:Un-coated
18
Degradesamplesundercondi<onsof45C,40%RH2Xsunlight(ordark)HemisphericalR
Silverdegradesdifferentlyinthedarkthanunderillumina<onReflectancechangesdifferentlywithAgthickness:thinfilminterference
0
10
20
30
40
50
60
70
80
90
100
250 750 1250 1750 2250
Refle
ctan
ce
Wavelength(nm)
5592-30-1,Cu(10)/Ag(100) 5592-30-1-1
5592-30-1-130hours
5592-30-1-160hours
5592-30-1-190hours
5592-30-1-1120hours
5592-30-1-2
5592-30-1-230hours
5592-30-1-260hours
5592-30-1-290hours
5592-30-1-3
5592-30-1-330hours
5592-30-1-360hours
5592-30-1-390hours
5592-30-1-3120hours
5592-30-1-4
5592-30-1-430hours
5592-30-1-460hours
5592-30-1-490hours
Substrate/Condi<on1-Glass/Light2-Glass/Dark3-3milPET/Light4-3milPET/Dark
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 19CSPProgramSummit2016
Agingofsimplifiedreflectorstacks
19
Iden<fywhatisbeingformedwithacombina<onof:• Variableanglespectroscopicellipsometry• RutherfordbackscaPering• X-rayphotoelectronspectroscopy• Specularityandroughnessmeasurements
Iden<fyrelevant<mescalesandplausiblekine<cmodels
• Forwardpredictchemicalchangesandcomparetoexp’tOngoingworktocorrelatechemistry/materialiden<tywithop<cal/mechanicalproper<estobuildpredic<vemodel
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 20CSPProgramSummit2016
Silvercorrosionkine'cs:Insightsfromtheliterature
20
• Quan<ta<veextrac<onofsilversulfida<onkine<cswithreac%on/advec%onmodelofdegrada<onofsilverindenudertube
FIG. 6. Photographs of Ag foils after 48 h exposure to corrosion-chamber air. Foil leading edge on the left. Flow rates through tube (cm3(STP) s-l): (a) 51.8, (b) 27.3, (c) 14.3,
(d) 7.67, (e) 3.83.
1187
2Ag+H2S+½O2->Ag2S+H2O
InputgasmixtureFixedflowrate
Exi<nggasmixture
Volpeetal.,CorrosionScience10,1179-1196(1989)
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 21CSPProgramSummit2016
Silvercorrosionkine'cs:Insightsfromtheliterature
21
• Quan<ta<veextrac<onofsilversulfida<onkine<cswithreac%on/advec%onmodelofdegrada<onofsilverindenudertube
Flowrates:3.83,7.67,14.3,27.3,51.8cm3(STP)/s
Volpeetal.,CorrosionScience10,1179-1196(1989)
2Ag+H2S+½O2->Ag2S+H2O
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 22CSPProgramSummit2016
Silvercorrosionkine'cs:Insightsfromtheliterature
• Lightdependence:arewephotolyzingO3?• ArewemakingAg2O,AgO
• Evolu<onofAgxOtoAg2O(aging)asmoreoxideforms?
• Areweformingaheterogeneousmixedlayerofoxidesandsulfides?Chlorides?
energy.gov/sunshotenergy.gov/sunshotCSPProgramSummit2016 23CSPProgramSummit2016
SummaryandOutlook
23
• Centralhypotheses• Fundamentalchemicaleventscanbemeasuredandmechanismsinferred
• Predic<vemodelscanbeconstructedbasedonthesemeasurements
• Measurementsshowcomplexdegrada<onkine<csevenforsimplifiedgeometriesandmaterials• Q:Isfullchemistryneeded?Canonegetawaywithasimplifiedpseudo-chemistry?
• APen<ontoroughening/lossofspecularity• Combinedop<cs/kine<cs/transportcodetobemade
availabletothecommunity