J Alzheimers Dis. Author manuscript; available in PMC 2018 Mar 21.
Published in final edited form as:
doi: 10.3233/JAD-130461
NIHMSID: NIHMS947764
The publisher's final edited version of this article is available at J Alzheimers Dis
See other articles in PMC that cite the published article.
Kyung Taek Cho, ORCID logo. The final perovskite film after the annealing process is called L-CFM/P (layered. In addition, the presence of the 2D perovskite in L-CFM/P can also be. Ho-Baillie, S.
Abstract
Previous preclinical studies have suggested a close relationship between cerebrovascular disease (CVD) and Alzheimer’s disease. However, a direct correlation between CVD and amyloid burden has not yet been shown in humans. If there is a relationship between CVD and amyloid burden, it is possible that the apolipoprotein E4 (APOE4) genotype may have an effect on this relationship because APOE4 is a risk factor for the development of AD. We therefore evaluated the effects of APOE4 on the relationship between white matter hyperintensities (WMH), a marker of CVD, and amyloid burden, measured by 11C-Pittsburgh compound B (PiB) PET. We recruited 53 patients with subcortical vascular cognitive impairments, who had both WMH on MRI and amyloid deposition assessed by PiB PET. Twenty-two of these patients were APOE4 carriers (41.5%). In the APOE4 non-carriers, a significant positive correlation was shown between the volume of WMH and PiB retention (β = 7.0 10−3, p = 0.034) while no significant correlation was found in APOE4 carriers (β = −9.0 × 10−3, p = 0.085). Statistical parametric mapping analyses in APOE4 non-carriers showed that WMH were associated with PiB retention in the bilateral medial occipitotemporal gyrus, cuneus, and superior cerebellum. Our results suggested that WMH are correlated with amyloid burden especially in the posterior brain regions in APOE4 non-carriers. However, this correlation was not observed in APOE4 carriers, perhaps because in these subjects the influence of APOE4 overrides the effect of CVD.
Keywords: Alzheimer’s disease, amyloid burden, apolipoprotein E4, cerebrovascular disease
INTRODUCTION
Subcortical vascular cognitive impairment (SVCI) refers to cognitive impairments due to cerebrovascular disease (CVD), which consists of subcortical vascular dementia (SVaD) and subcortical vascular mild cognitive impairment (svMCI) []. Pathological studies, however, demonstrated that patients who had been clinically diagnosed with SVaD often proved to have comorbid Alzheimer’s disease (AD) pathologies [–]. A study by our group using Pittsburgh compound-B (PiB) PET, a sensitive method to detect amyloid plaque burden during life [], also showed that over 30% of SVaD patients turned out to be PiB-positive [].
In epidemiologic studies, there is increasing evidence that CVD and AD dementia are strongly associated [–]. Some preclinical studies also suggest that CVD may directly induce amyloid burden []. Therefore, it has been suggested that CVD may accelerate the development of AD []. There are two possible mechanisms that could explain the growing evidence linking CVD and AD dementia. The first is that CVD accelerates the deposition of brain amyloid, which leads to AD dementia. The second possible interpretation is that CVD produces brain damage, which would reduce cognitive reserve and thus make the subjects more susceptible to the effects of AD pathology. One recent study of human subjects was unable to demonstrate any direct correlation between CVD and amyloid burden []. This result would be most consistent with the second mechanism (described above) that CVD lowers cognitive reserve without having a direct effect on development of brain amyloid pathology. However, this previous study did not consider the effect of the apolipoprotein E ε4 (APOE4) genotype on the relationship between CVD and brain amyloid. The APOE4 genotype is considered a major risk factor for AD and is thought to lower the age of onset for the development of AD and accelerate the aggregation of amyloid plaques []. Some studies also showed that APOE4 is associated with CVD, although their results were inconsistent [–]. Thus, it is possible that the APOE4 genotype might have an effect on the relationship between CVD and amyloid burden.
The current study aimed to examine the effects of APOE4 on the relationship between CVD, which was quantified as white matter hyperintensities (WMH) on fluid-attenuated inversion recovery (FLAIR) MRI, and brain amyloid, which was measured by PiB PET imaging. We evaluated the relationship between WMH and amyloid burden in APOE4 carriers and non-carriers, respectively. We hypothesized that there may be a correlation between the two pathologies affected by the APOE4 genotype status.
MATERIALS AND METHODS
Participants
A total of 136 SVCI patients were consecutively recruited at Samsung Medical Center from September 2008 to August 2011. Patients with SVaD met the diagnostic criteria for vascular dementia as determined by the Diagnostic and Statistical Manual of Mental Disorders-Fourth Edition (DSM-IV). DSM-IV criteria include the presence of focal signs suggestive of CVD; we defined the presence of focal signs as at least two focal neurologic signs out of corticobular, corticospinal, extrapyramidal signs and gait abnormalities. All SVaD patients had a significant ischemia on their MRI scans, which was defined as a volume of WMH ≥15 cm3. This cutoff value of 15 cm3 was the smallest volume of WMH in patients who met grade 3 modified Fazekas ischemia criteria. A prior study suggested that WMH volume measurement might be more sensitive than visual scores []. Patients with svMCI were diagnosed using the Petersen criteria [] with the following modifications: 1) there was subjective cognitive complaint by the patient or his/her caregiver; 2) normal Activity of Daily Living (ADL) score by clinically and ADL scale (Seoul Instrumental ADL score below seven, which is a modified version of Lawton’s 1969 Instrumental ADL) [21, ]; 3) the patient showed an objective memory decline below the 16th percentile on neuropsychological tests [, , ]; 4) not having dementia; and 5) presence of a subcortical vascular feature defined as both focal neurological symptom or sign and the significant ischemia on MRI, as in the SVaD. We excluded patients with other structural lesions on brain MRI such as territorial infarction, intracranial hemorrhage, traumatic brain injury, hydrocephalus, or WMH associated with radiation, multiple sclerosis, or vasculitis. The number of patients with svMCI and SVaD were 59 (43.4%) and 77 (56.6%), respectively.
Patients with SVCI were classified as PiB-positive (PiB+) or PiB-negative (PiB−) if they had a global PiB retention ratio greater than or less than 1.5, respectively []. Among the 136 patients with SVCI, 53 (39.0%) were positive for PiB retention, while 83 (61.0%) were negative. Since our aim was to investigate the relationship between the two pathologies, we considered that subjects that have only one of the two pathologies may be inappropriate for our study goal. Therefore, we only included, from our SVCI cohort, the SVCI patients with PiB-positive scan. Among the 53 patients who were PiB+, 58.5% were APOE4 non-carriers (n = 31: ε2/ε3, n = 6; ε3/ε3, n = 25), and 41.5% were APOE4 carriers (n = 22: ε3/ε4, n = 18; ε4/ε4, n = 4).
All patients completed a clinical interview and neurological examination, blood tests, and APOE genotype as described previously [], and MRI scans at Samsung Medical Center using the same scanner. All patients also completed a three-step diagnostic process. First, patients completed a medical interview conducted by an experienced neurologist, who obtained a medical history, a history of cognitive, behavioral, and functional impairments, and performed neurological examinations, including the Mini-Mental Status Exam (MMSE), Clinical Dementia Rating Sum of Boxes (CDR-SOB), and Geriatric Depression Scale (GDS). The interview also included items to assess patients’ abilities to engage in activities of daily living. Second, the neuropsychology team performed a number of neuropsychological tests and conducted a clinical interview for cognitive, behavioral, and functional impairments using semi-structured questionnaires. The scales for Neuropsychiatric Inventory and Activities of Daily Living (ADL) scales were completed. The ADL scale used in this study was Seoul Instrumental ADL score, which is a modified version of Lawton’s Instrumental ADL [21, ]. Third, patients were diagnosed based on the results of all the diagnostic tests (neuropsychological reports, blood tests, and MRIs). Blood tests included a complete blood count, blood chemistry test, vitamin B12/folate measure, syphilis serology, thyroid functioning tests, and APOE genotyping. We obtained a written consent from each patient and the Institutional Review Board of the Samsung Medical Center approved the study protocol.
Neuropsychological tests
All patients underwent neuropsychological testing using the Seoul Neuropsychological Screening Battery (SNSB) [, ]. This battery contains assessments of attention, language abilities, praxis, four elements of Gerstmann syndrome, visuospatial functioning, verbal and visual memory, and frontal/executive functioning. Among these subtests, the quantitatively scorable tests, including digit span (forward and backward), the Korean version of the Boston Naming Test (K-BNT) [], the Rey-Osterrieth Complex Figure Test (RCFT; copying, immediate and 20-min delayed recall, and recognition), Seoul Verbal Learning Test (SVLT; three learning-free recall trials of 12 words, a 20-minute delayed recall trial for these 12 items, and a recognition test), phonemic and semantic Controlled Oral Word Association Test (COWAT), and a Stroop Test (word and color reading of 112 items during a 2-min period) were used in current study.
MR imaging and analysis
T2, T1, three-dimensional (3D) FLAIR, and T2 fast field echo (FFE) images were acquired for all 53 subjects at Samsung Medical Center using the same 3.0 T MRI scanner (Philips 3.0T Achieva). The MRI scanning protocol was the same as described previously [].
Measurement of regional WMH volume
We quantified WMH volume (in cm3) on FLAIR images using an automated method as previously described [28]. First, we extracted the WMH candidate regions on FLAIR images applying classification method and morphological operation to T1-weighted images. Second, in order to extract WMH, a threshold method was applied to the FLAIR images within the WMH candidate regions. Even though the threshold value was selected considering the range of image intensities, segmented results could contain false positive or false negative regions depending on the extent of WMH. If the results contained an error, the threshold value was reselected through visual inspection by two raters, and they reached a consensus in the case of discrepancy. The rate of agreement between two neurologists was 92.3%.
[11C] PiB-PET imaging
All patients completed a standardized [11C] PiB-PET scan spanning the entire brain at Samsung Medical Center or Asan Medical Center using a Discovery STe PET/CT scanner (GE Medical Systems, Milwaukee, WI) in 3-dimensional scanning mode that examined 35 slices, each 4.25-mm thick. First, the 11C-PiB was injected into an antecubital vein as a bolus with a mean dose of 420 MBq (range 259–550 MBq). Sixty minutes after the injection, a CT scan was performed for attenuation correction. Afterwards, the 30-minute emission static PET scan was then initiated [].
Data analysis of [11C] PiB-PET images
PiB PET images were co-registered to each individual’s MRIs, which were normalized to a T1-weighted MRI template. Using these parameters, MRI co-registered PiB PET images were normalized to the MRI template. The quantitative regional values of PiB retention on the spatially normalized PiB images were obtained by an automated volume of interest (VOI) analysis tool using the automated anatomical labeling atlas. Data processing was performed using SPM version 8 (SPM8) through Matlab 6.5 (Mathworks, Natick, MA, USA).
To measure PiB retention, we used the cerebral cortical region to cerebellum uptake ratio which is identical to the standardized uptake value ratios (SUVRs). The cerebellum was used as a reference region as it did not show group differences. We selected 28 cortical VOIs from left as well as right hemispheres using the automated anatomical labeling atlas. The cerebral cortical VOIs which were chosen for this study consisted of the following areas: bilateral frontal (superior and middle frontal gyri, medial part of superior frontal gyrus, opercular part of inferior frontal gyrus, triangular part of inferior frontal gyrus, supplementary motor area, orbital part of superior, middle, and inferior orbital frontal gyri, rectus and olfactory cortex), posterior cingulate gyri, parietal (superior and inferior parietal, supramarginal and angular gyri, and precuneus), lateral temporal (superior, middle and inferior temporal gyri, and heschl gyri), and occipital (superior, middle, and inferior occipital gyri, cuneus, calcarine fissure, and lingual and fusiform gyri). Regional cerebral cortical SUVRs were calculated by dividing each cortical VOI’s SUV by the mean SUV of the cerebellar cortex (cerebellum crus1 and crus2). The global PiB retention ratio was calculated from the volume-weighted averages of the SUVRs of the bilateral cerebral cortical VOIs. We defined the PiB retention ratio as a continuous variable.
Statistical analysis
Comparisons of demographic and clinical data between APOE4 non-carriers and carriers were conducted using Student’s t test for normally distributed variables, and the Mann-Whitney U test for non-normally distributed variables. Categorical variables were evaluated using a χ2 test. In order to examine the relationship between WMH and PiB retention ratio, multiple linear regression analyses were performed using WMH as an independent factor and the global PiB retention ratio as a dependent variable with adjustment for age and gender. Assumptions of residual about normality, homoscedasticity, and independence were confirmed. Statistical significance was set at p < 0.05. Statistical analyses were conducted using PASW Statistics 18 (SPSS Inc, Chicago, IL, USA) software.
In order to examine the relationship between WMH and PiB retention and to determine the regions where this relationship was significant, a voxel-based statistical analysis of the PiB images was performed using the Statistical Parametric Mapping program, version 8 (SPM8), and Matlab 6.5 for Windows (Math Works, Natick, MA, USA). An SPM regression analysis was performed without global normalization, since the 11C-PiB PET images had been normalized to the cerebellar ROI PiB binding. Multivariate regression analysis was performed using WMH as a predictor and the PiB retention ratio at each voxel as a dependent variable after adjusting for age and gender. Red or yellow colored area showed the statistically significant region with increased global PiB retention as the volume of WMH increased with adjustment for age and gender. We defined statistical significance as p < 0.001 (uncorrected for multiple comparisons).
We generated group-specific gray matter mask using T1-weighted images in order to exclude non-gray matter areas from the PiB images. We used classification method and chose the threshold of 0.5 to binarize gray matter probability map for the mask. Results were overlaid to a T1 weighted MRI of a representative male individual template provided by SPM.
RESULTS
Baseline characteristics of the subjects
Demographic characteristics and imaging findings of the 53 patients are presented in Table 1. There were no differences in demographics or severity of cognitive impairment between APOE4 non-carriers and carriers. There were also no differences in the median volume of WMH or mean global PiB SUVR between the two APOE groups.
Table 1
Demographic characteristics and imaging findings in study population
APOE4 Non-carriers | APOE4 Carriers | p value‖ | |
---|---|---|---|
Number | 31 | 22 | |
Mild cognitive impairment* | 11 (35.5%) | 8 (36.4%) | >0.999 |
Age† | 78.39 ± 5.14 | 76.41 ± 5.83 | 0.197 |
Male, N (%)* | 13 (41.9%) | 6 (27.3%) | 0.420 |
Education (year)† | 9.65 ± 6.21 | 9.91 ± 5.15 | 0.871 |
K-MMSE† | 20.87 ± 5.78 | 22.59 ± 5.70 | 0.288 |
CDR SOB‡ | 3.50 (1.50–7.00) | 3.50(1.50–6.00) | 0.950 |
Hypertension* | 17 (54.8%) | 18 (81.8%) | 0.080 |
Diabetes mellitus* | 7 (22.6%) | 6 (27.3%) | 0.946 |
Dyslipidemia* | 6 (19.4%) | 7 (31.8%) | 0.475 |
Coronary artery disease§ | 5 (16.1%) | 3 (13.6%) | >0.999 |
Stroke history§ | 5 (16.1%) | 3 (13.6%) | >0.999 |
WMH volume (cm3)‡ | 36.02 (24.34–49.93) | 33.77 (21.94–45.94) | 0.626 |
Microbleed prevalence* | 18(58.1%) | 11 (50.0%) | 0.561 |
Microbleed number‡ | 1.0 (0–4.0) | 0.5 (0–4.0) | 0.812 |
Global PiB retention ratio† | 2.07 ± 0.38 | 2.10 ± 0.38 | 0.760 |
Data are presented as mean ± standard deviation for normally distributed and median (25th–75th percentile, interquartile range) for non-normally distributed variables.
*χ2 test with Yates’ continuity correction was used.
†Student’s t test was used for normally distributed variables.
‡Mann-Whitney U test was used for non-normally distributed variables.
‖p-value for the comparison of the APOE4 non-carrier group and APOE4 carrier group. K-MMSE, Korean version of the Mini-Mental State Examination; CDR SOB, Clinical Dementia Rating Sum of Box; WMH, white matter hyperintensities.
Correlations between WMH and global PiB retention ratio in APOE4 non-carriers and carriers
In the whole PiB+cohort, there was no correlation between WMH volume and cortical PiB retention (unstandardized coefficient β = 4.0 × 10−3, SE (β) = 3.0 × 10−3, p = 0.175).
In the APOE4 non-carriers, a significant positive correlation was shown between the volume of WMH and cortical PiB retention (β = 7.0 × 10−3, SE(β) = 3.0 × 10−3, p = 0.034). However, no positive correlation was found in APOE4 allele carriers (β = −9.0 × 10−3, SE (β) = 5.0 × 10−3, p = 0.085) (Fig. 1).
Scatter plot graph between the volume of white matter hyperintensities (WMH) (cm3) and global PiB retention ratio. Multiple linear regression analyses were performed using WMH as an independent factor and the global PiB retention ratio as a dependent variable with adjustment for age and gender. A) Positive correlation with statistical significance was shown between WMH volumes and PiB retention areas in APOE4 non-carrier (β = 7.0 × 10−3, SE(β) = 3.0 × 10−3, p = 0.034). B) Significant correlation was not shown between WMH volumes and PiB retention in APOE4 carrier (β = −9.0 × 10−3, SE(β) = 5.0 × 10−3, p = 0.085).
To evaluate the effect modification of APOE4 status and WMH volume on amyloid burden, multiple linear regressions were performed in all APOE4 carriers and non-carriers (n = 53 for total APOE4 carriers and non-carriers). We used age, gender, WMH, APOE4 status, and the interaction terms (APOE4 status*WMH volume) as independent variables, and PiB retention ratio as the dependent variable. There tended to be interactions between APOE4 status and WMH volume (p = 0.090).
Regional correlation between WMH and cortical PiB retention in APOE4 non-carriers and carriers
In APOE4 non-carriers, WMH volume was positively correlated with PiB retention in the bilateral medial occipitotemporal gyrus, cuneus, and the superior part of the cerebellum such as culmen, declive, anterior quadrangular lobule (uncorrected p < 0.001) (Fig. 2). However, after false discovery rate (FDR) correction, the statistical significance disappeared in these regions.
SPM multiple regression analysis of PiB retention in the APOE4 non-carriers. A) t value map: The red or yellow colored regions represents the significant positive correlation between white matter hyperintensities volumes and PiB retention ratio with adjustment for age and gender (uncorrected p < 0.001). B) Beta map: Strength of associations is illustrated with color-labeled standardized beta-values.
In APOE4 carriers, there were no regions showing the significant correlation between WMH volumes and PiB retention ratio. However, WMH volumes tend to have a negative correlation with PiB retention ratio in the right dorsolateral frontal, precuneus, and bilateral parietotemporal areas (uncorrected p < 0.01) (Fig. 3).
SPM multiple regression analysis of PiB retention in the APOE4 carriers. There were no regions showing the significant correlation between WMH volumes and PiB retention ratio. A) t value map: The blue or sky blue colored regions represent the negative correlation between WMH and PiB retention ratio with adjustment for age and gender (uncorrected p < 0.01). B) Beta map: Strength of associations is illustrated with color-labeled standardized beta-values.
DISCUSSION
Our major findings were as follows. First, there was a significant positive relationship between the volume of WMH and PiB retention in the APOE4 non-carriers, whereas no significant analogous relationship was seen in APOE4 carriers. Second, the brain regions where WMH correlated with PiB retention were in the temporal, occipital, and superior cerebellar regions. Our findings suggest that WMH is associated with amyloid burden, especially in the posterior brain regions of APOE4 non-carriers with PiB (+) SVCI. These findings raise the possibility that APOE4 interacts with the pathogenesis of amyloid deposition in the presence of CVD. Possible mechanisms are discussed below.
Our first major finding of a positive correlation between the volume of WMH and PiB retention ratio in APOE4 non-carriers is consistent with many epidemiologic studies demonstrating an association between CVD and AD pathology [–, , ]. Furthermore, a recent preclinical study showed the direct relationship between ischemic lesion and amyloid burden []. A previous study also suggested that patients with stroke had increased PiB retention in the peri-infarct region []. In contrast, one PiB PET study, based on healthy subjects with normal cognition, showed there was no correlation between the two pathologies [], but this study did not factor in the effect of the APOE4 genotype. Indeed, in the present study, we found no correlation between WMH and amyloid burden in the carriers and non-carriers combined. Another reason for the discrepancy between previous findings and our current findings may be the difference in the subject population. Our study included MCI and dementia patients with greater amounts of WMH and brain amyloid in comparison with the previous asymptomatic community subjects [] with lower amounts of WMH and brain amyloid. Our finding might be supported by a more recent study showing that baseline WMH was positively correlated with the progression of amyloid burden over 2 years []. The mechanisms underpinning the positive correlation between WMH and PiB retention remain unclear. However, it is possible that these two factors interact with each other (interactive hypothesis). One hypothesis is that CVD, reflected by WMH can block the clearance of amyloid via perivascular lymphatic drainage [–]. Alternatively, amyloid deposition in small or medium sized arterioles can lead to amyloid angiopathy, which may also result in severe WMH [, ]. However, since both AD and vascular pathologies usually occur in the elderly and share risk factors such as hypertension, diabetes, and hyperlipidemia, it is also possible that both pathologies can occur independently by chance [, , ].
Our second major finding was that WMH were associated with amyloid burden predominantly in the temporal, occipital and superior cerebellar regions, which are not typically affected by amyloid burden in AD []. This finding is similar to a recent report that amyloid deposition occurred predominantly in the occipital areas and cerebellum in patients with carotid stenosis []. Our finding is also consistent with another previous study showing that baseline WMH was associated with progression of amyloid burden in the parieto-occipital region []. The reason why WMH were correlated with amyloid burden in the posterior brain region in APOE4 non-carriers remains unclear. We suggest that there are at least two possibilities to explain the relationship between WMH and amyloid: (1) the distribution corresponds to the posterior circulation, which is supplied by vertebrobasilar system; [] or (2) it is related to the topography of cerebral amyloid angiopathy (CAA). The posterior circulation may be vulnerable to injury and dysfunction of the endothelium leading to blood-brain barrier (BBB) disruption []. For example, patients with hypertensive encephalopathy have vasculopathic changes predominantly in the posterior circulation areas. Furthermore, emerging evidence suggests that BBB disruption may contribute to the development of AD [, –], possibly through increased blockage of amyloid clearance. Therefore, it is possible that CVD may result in preferential BBB disruption in the posterior circulation, which may induce increased amyloid deposition in these areas. Alternatively, the association between WMH and amyloid burden might be related to CAA, which involves predominantly the occipital and temporal regions [, , ].
We found no correlation between WMH and PiB retention in APOE4 carriers, similar to a previous report that did not account for APOE4 []. Although animal studies have shown that APOE4 is associated with enhanced amyloid-β aggregation [, ] and reduced amyloid clearance [, , ], resulting in amyloid-β deposition, the mechanism by which APOE4 promotes earlier age of amyloid deposition in humans is not known. Amyloid imaging studies have repeatedly shown that age and APOE4 status are the primary predictors of brain amyloid load. Therefore, we suggest that the powerful effects of APOE4 on brain amyloid override the effects of CVD (WMH) on brain amyloid in APOE4 positive subjects. Alternatively, it might be related to the explanation that SVCI patients with severe WMH have less amyloid burden than SVCI patients with moderate WMH because both WMH and amyloid burden causes cognitive impairments. Indeed, there was a trend displaying negative correlation between WMH and amyloid burdens in APOE4 carriers.
There are some limitations to our study. First, we were unable to test the hypothesis of a causal relationship between ischemia and amyloid burden because this was a cross-sectional study. Longitudinal studies with serial MRI and PiB PET would be able to examine this issue. Second, all of the patients in this study had cognitive impairment, which may limit the generalizability of the results. However, since both CVD and amyloid pathology are related to cognitive impairment, our population is representative of an important clinical manifestation of these processes. Third, we were unable to differentiate the parenchymal amyloid from vascular amyloid using PiB-PET. However, in this study, there were no subjects who met the clinical criteria for CAA [] or showed restricted lobar microbleeds. Finally, after corrections for multiple comparisons with FDR, significant associations in the SPM analysis disappeared. However, to reduce the chances of missing important associations during the early stage of analysis, we used a statistical significance level of 0.001 without correction for multiple comparisons.
Nevertheless, this study is noteworthy because it is the first human study demonstrating the correlation of severity of small vessel related ischemia and amyloid burden in APOE4 non-carriers, especially in the posterior cortical regions. Further investigation on the interaction of CVD and amyloid deposition may provide clues to treatment or prevention strategies for dementia.
Acknowledgments
This study was supported by Basic Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF-2013R1A1A2065365), the Korean Healthcare Technology R&D Project Ministry for Health & Welfare Affairs, Republic of Korea (HI10C2020 & HIC120713), by the KOSEF NRL program grant (MEST; 2011-0028333), by Samsung Medical Center (CRL-108011&CRS110-14-1), and by the Converging Research Center Program through the Ministry of Science, ICT and Future Planning, Korea (2013K000338).
Footnotes
Handling Associate Editor: Josephine Barnes
Authors’ disclosures available online (http://www.j-alz.com/disclosures/view.php?id=2075).
References
1. Seo SW, Ahn J, Yoon U, Im K, Lee JM, Tae Kim S, Ahn HJ, Chin J, Jeong Y, Na DL. Cortical thinning in vascular mild cognitive impairment and vascular dementia of subcortical type. J Neuroimaging. 2010;20:37–45. [PubMed] [Google Scholar]
2. Schneider JA, Arvanitakis Z, Bang W, Bennett DA. Mixed brain pathologies account for most dementia cases in community-dwelling older persons. Neurology. 2007;69:2197–2204. [PubMed] [Google Scholar]
3. Chui HC, Zarow C, Mack WJ, Ellis WG, Zheng L, Jagust WJ, Mungas D, Reed BR, Kramer JH, Decarli CC, Weiner MW, Vinters HV. Cognitive impact of subcortical vascular and Alzheimer’s disease pathology. Ann Neurol. 2006;60:677–687.[PMC free article] [PubMed] [Google Scholar]
4. Jagust WJ, Zheng L, Harvey DJ, Mack WJ, Vinters HV, Weiner MW, Ellis WG, Zarow C, Mungas D, Reed BR, Kramer JH, Schuff N, DeCarli C, Chui HC. Neuropathological basis of magnetic resonance images in aging and dementia. Ann Neurol. 2008;63:72–80.[PMC free article] [PubMed] [Google Scholar]
5. Klunk WE, Engler H, Nordberg A, Wang Y, Blomqvist G, Holt DP, Bergstrom M, Savitcheva I, Huang GF, Estrada S, Ausen B, Debnath ML, Barletta J, Price JC, Sandell J, Lopresti BJ, Wall A, Koivisto P, Antoni G, Mathis CA, Langstrom B. Imaging brain amyloid in Alzheimer’s disease with Pittsburgh Compound-B. Ann Neurol. 2004;55:306–319. [PubMed] [Google Scholar]
6. Lee JH, Kim SH, Kim GH, Seo SW, Park HK, Oh SJ, Kim JS, Cheong HK, Na DL. Identification of pure subcortical vascular dementia using 11C-Pittsburgh compound B. Neurology. 2011;77:18–25. [PubMed] [Google Scholar]
7. Hofman A, Ott A, Breteler MM, Bots ML, Slooter AJ, van Harskamp F, van Duijn CN, Van Broeckhoven C, Grobbee DE. Atherosclerosis, apolipoprotein E, and prevalence of dementia and Alzheimer’s disease in the Rotterdam Study. Lancet. 1997;349:151–154. [PubMed] [Google Scholar]
8. Roher AE, Tyas SL, Maarouf CL, Daugs ID, Kokjohn TA, Emmerling MR, Garami Z, Belohlavek M, Sabbagh MN, Sue LI, Beach TG. Intracranial atherosclerosis as a contributing factor to Alzheimer’s disease dementia. Alzheimers Dement. 2011;7:436–444.[PMC free article] [PubMed] [Google Scholar]
9. Shah NS, Vidal JS, Masaki K, Petrovitch H, Ross GW, Tilley C, DeMattos RB, Tracy RP, White LR, Launer LJ. Midlife blood pressure, plasma beta-amyloid, and the risk for Alzheimer disease: The Honolulu Asia Aging Study. Hypertension. 2012;59:780–786.[PMC free article] [PubMed] [Google Scholar]
10. Schmidt R, Ropele S, Enzinger C, Petrovic K, Smith S, Schmidt H, Matthews PM, Fazekas F. White matter lesion progression, brain atrophy, and cognitive decline: The Austrian stroke prevention study. Ann Neurol. 2005;58:610–616. [PubMed] [Google Scholar]
11. Garcia-Alloza M, Gregory J, Kuchibhotla KV, Fine S, Wei Y, Ayata C, Frosch MP, Greenberg SM, Bacskai BJ. Cerebrovascular lesions induce transient beta-amyloid deposition. Brain. 2011;134:3697–3707.[PMC free article] [PubMed] [Google Scholar]
12. Petrovitch H, White LR, Izmirilian G, Ross GW, Havlik RJ, Markesbery W, Nelson J, Davis DG, Hardman J, Foley DJ, Launer LJ. Midlife blood pressure and neuritic plaques, neurofibrillary tangles, and brain weight at death: The HAAS. Honolulu-Asia aging Study. Neurobiol Aging. 2000;21:57–62. [PubMed] [Google Scholar]
13. Marchant NL, Reed BR, DeCarli CS, Madison CM, Weiner MW, Chui HC, Jagust WJ. Cerebrovascular disease. beta-amyloid, and cognition in aging. Neurobiol Aging. 2012;33:1006 e1025–e1036.[PMC free article] [PubMed] [Google Scholar]
14. Castellano JM, Kim J, Stewart FR, Jiang H, DeMattos RB, Patterson BW, Fagan AM, Morris JC, Mawuenyega KG, Cruchaga C, Goate AM, Bales KR, Paul SM, Bateman RJ, Holtzman DM. Human apoE isoforms differentially regulate brain amyloid-beta peptide clearance. Sci Transl Med. 2011;3:89ra57.[PMC free article] [PubMed] [Google Scholar]
15. Altamura C, Squitti R, Pasqualetti P, Tibuzzi F, Silvestrini M, Ventriglia MC, Cassetta E, Rossini PM, Vernieri F. What is the relationship among atherosclerosis markers, apolipoprotein E polymorphism and dementia? Eur J Neurol. 2007;14:679–682. [PubMed] [Google Scholar]
16. Lahoz C, Schaefer EJ, Cupples LA, Wilson PW, Levy D, Osgood D, Parpos S, Pedro-Botet J, Daly JA, Ordovas JM. Apolipoprotein E genotype and cardiovascular disease in the Framingham Heart Study. Atherosclerosis. 2001;154:529–537. [PubMed] [Google Scholar]
17. Graner M, Kahri J, Varpula M, Salonen RM, Nyyssonen K, Jauhiainen M, Nieminen MS, Syvanne M, Taskinen MR. Apolipoprotein E polymorphism is associated with both carotid and coronary atherosclerosis in patients with coronary artery disease. Nutr Metab Cardiovasc Dis. 2008;18:271–277. [PubMed] [Google Scholar]
18. Kim HJ, Ye BS, Yoon CW, Cho H, Noh Y, Kim GH, Choi YS, Kim JH, Jeon S, Lee JM, Kim JS, Choe YS, Lee KH, Kim ST, Kim C, Kang DR, Ki CS, Lee JH, Werring DJ, Weiner MW, Na DL, Seo SW. Effects of APOE epsilon4 on brain amyloid, lacunar infarcts, and white matter lesions: A study among patients with subcortical vascular cognitive impairment. Neurobiol Aging. 2013;34:2482–2487. [PubMed] [Google Scholar]
19. van Straaten EC, Fazekas F, Rostrup E, Scheltens P, Schmidt R, Pantoni L, Inzitari D, Waldemar G, Erkinjuntti T, Mantyla R, Wahlund LO, Barkhof F. Impact of white matter hyperintensities scoring method on correlations with clinical data: The LADIS study. Stroke. 2006;37:836–840. [PubMed] [Google Scholar]
20. Petersen RC. Mild cognitive impairment as a diagnostic entity. J Intern Med. 2004;256:183–194. [PubMed] [Google Scholar]
21. Ku HM, Kim JH, Kwon EJ, Kim SH, Lee HS, Ko HJ, Jo S, Kim DK. A study on the reliability and validity of Seoul-Instrumental Activities of Daily Living(S-IADL) J Korean Neuropsychiatr Assoc. 2004;43:189–199.[Google Scholar]
22. Lawton MP, Brody EM. Assessment of older people: Self-maintaining and instrumental activities of daily living. Gerontologist. 1969;9:179–186. [PubMed] [Google Scholar]
23. Chin J, Seo SW, Kim SH, Park A, Ahn HJ, Lee BH, Kang SJ, Na DL. Neurobehavioral dysfunction in patients with subcortical vascular mild cognitive impairment and subcortical vascular dementia. Clin Neuropsychol. 2012;26:224–238. [PubMed] [Google Scholar]
24. Seo SW, Lee JM, Im K, Park JS, Kim SH, Kim ST, Ahn JH, Kim MJ, Kim GH, Kim JH, Roh JH, Cheong HK, Na DL. Cardiovascular risk factors cause cortical thinning in cognitively impaired patients: Relationships among cardiovascular risk factors, white matter hyperintensities, and cortical atrophy. Alzheimer Dis Assoc Disord. 2012;26:106–112. [PubMed] [Google Scholar]
25. Ahn HJ, Chin J, Park A, Lee BH, Suh MK, Seo SW, Na DL. Seoul Neuropsychological Screening Battery-dementia version (SNSB-D): A useful tool for assessing and monitoring cognitive impairments in dementia patients. J Korean Med Sci. 2010;25:1071–1076.[PMC free article] [PubMed] [Google Scholar]
26. Ahn HJ, Seo SW, Chin J, Suh MK, Lee BH, Kim ST, Im K, Lee JM, Lee JH, Heilman KM, Na DL. The cortical neuroanatomy of neuropsychological deficits in mild cognitive impairment and Alzheimer’s disease: A surface-based morphometric analysis. Neuropsychologia. 2011;49:3931–3945. [PubMed] [Google Scholar]
27. Kim H, Na DL. Normative data on the Korean version of the Boston Naming Test. J Clin Exp Neuropsychol. 1999;21:127–133. [PubMed] [Google Scholar]
28. Jeon S, Yoon U, Park JS, Seo SW, Kim JH, Kim ST, Kim SI, Na DL, Lee JM. Fully automated pipeline for quantification and localization of white matter hyperintensity in brain magnetic resonance image. Int J Imaging Syst Technol. 2011;21:193–200.[Google Scholar]
29. Pappolla MA, Bryant-Thomas TK, Herbert D, Pacheco J, Fabra Garcia M, Manjon M, Girones X, Henry TL, Matsubara E, Zambon D, Wolozin B, Sano M, Cruz-Sanchez FF, Thal LJ, Petanceska SS, Refolo LM. Mild hypercholesterolemia is an early risk factor for the development of Alzheimer amyloid pathology. Neurology. 2003;61:199–205. [PubMed] [Google Scholar]
30. Ly JV, Rowe CC, Villemagne VL, Zavala JA, Ma H, Sahathevan R, O’Keefe G, Gong SJ, Gunawan R, Churilov L, Saunder T, Ackerman U, Tochon-Danguy H, Donnan GA. Subacute ischemic stroke is associated with focal 11C PiB positron emission tomography retention but not with global neocortical Abeta deposition. Stroke. 2012;43:1341–1346. [PubMed] [Google Scholar]
31. Grimmer T, Faust M, Auer F, Alexopoulos P, Forstl H, Henriksen G, Perneczky R, Sorg C, Yousefi BH, Drzezga A, Kurz A. White matter hyperintensities predict amyloid increase in Alzheimer’s disease. Neurobiol Aging. 2012;33:2766–2773. [PubMed] [Google Scholar]
32. Weller RO, Subash M, Preston SD, Mazanti I, Carare RO. Perivascular drainage of amyloid-beta peptides from the brain and its failure in cerebral amyloid angiopathy and Alzheimer’s disease. Brain Pathol. 2008;18:253–266. [PubMed] [Google Scholar]
33. Zlokovic BV. Neurovascular pathways to neurodegeneration in Alzheimer’s disease and other disorders. Nat Rev Neurosci. 2011;12:723–738.[PMC free article] [PubMed] [Google Scholar]
34. Viswanathan A, Greenberg SM. Cerebral amyloid angiopathy in the elderly. Ann Neurol. 2011;70:871–880.[PMC free article] [PubMed] [Google Scholar]
35. Gurol ME, Viswanathan A, Gidicsin C, Hedden T, Martinez-Ramirez S, Dumas A, Vashkevich A, Ayres AM, Auriel E, van Etten E, Becker A, Carmasin J, Schwab K, Rosand J, Johnson KA, Greenberg SM. Cerebral amyloid angiopathy burden associated with leukoaraiosis: A positron emission tomography/magnetic resonance imaging study. Ann Neurol. 2013;73:529–536.[PMC free article] [PubMed] [Google Scholar]
36. de la Monte SM, Wands JR. Alzheimer’s disease is type 3 diabetes-evidence reviewed. J Diabetes Sci Technol. 2008;2:1101–1113.[PMC free article] [PubMed] [Google Scholar]
37. Joachim CL, Morris JH, Selkoe DJ. Diffuse senile plaques occur commonly in the cerebellum in Alzheimer’s disease. Am J Pathol. 1989;135:309–319.[PMC free article] [PubMed] [Google Scholar]
38. Huang KL, Lin KJ, Ho MY, Chang YJ, Chang CH, Wey SP, Hsieh CJ, Yen TC, Hsiao IT, Lee TH. Amyloid deposition after cerebral hypoperfusion: Evidenced on [(18)F]AV-45 positron emission tomography. J Neurol Sci. 2012;319:124–129. [PubMed] [Google Scholar]
39. Caplan L. Posterior circulation ischemia: Then, now, and tomorrow. The Thomas Willis Lecture-2000. Stroke. 2000;31:2011–2023. [PubMed] [Google Scholar]
40. Feske SK. Posterior reversible encephalopathy syndrome: A review. Semin Neurol. 2011;31:202–215. [PubMed] [Google Scholar]
41. Bell RD, Zlokovic BV. Neurovascular mechanisms and blood-brain barrier disorder in Alzheimer’s disease. Acta Neuropathol. 2009;118:103–113.[PMC free article] [PubMed] [Google Scholar]
42. Deane R, Du Yan S, Submamaryan RK, LaRue B, Jovanovic S, Hogg E, Welch D, Manness L, Lin C, Yu J, Zhu H, Ghiso J, Frangione B, Stern A, Schmidt AM, Armstrong DL, Arnold B, Liliensiek B, Nawroth P, Hofman F, Kindy M, Stern D, Zlokovic B. RAGE mediates amyloid-beta peptide transport across the blood-brain barrier and accumulation in brain. Nat Med. 2003;9:907–913. [PubMed] [Google Scholar]
43. Ujiie M, Dickstein DL, Carlow DA, Jefferies WA. Blood-brain barrier permeability precedes senile plaque formation in an Alzheimer disease model. Microcirculation. 2003;10:463–470. [PubMed] [Google Scholar]
44. Zlokovic BV, Ghiso J, Mackic JB, McComb JG, Weiss MH, Frangione B. Blood-brain barrier transport of circulating Alzheimer’s amyloid beta. Biochem Biophys Res Commun. 1993;197:1034–1040. [PubMed] [Google Scholar]
45. Johnson KA, Gregas M, Becker JA, Kinnecom C, Salat DH, Moran EK, Smith EE, Rosand J, Rentz DM, Klunk WE, Mathis CA, Price JC, Dekosky ST, Fischman AJ, Greenberg SM. Imaging of amyloid burden and distribution in cerebral amyloid angiopathy. Ann Neurol. 2007;62:229–234. [PubMed] [Google Scholar]
46. Weller RO, Nicoll JA. Cerebral amyloid angiopathy: Both viper and maggot in the brain. Ann Neurol. 2005;58:348–350. [PubMed] [Google Scholar]
47. Bales KR, Liu F, Wu S, Lin S, Koger D, DeLong C, Hansen JC, Sullivan PM, Paul SM. Human APOE isoform-dependent effects on brain beta-amyloid levels in PDAPP transgenic mice. J Neurosci. 2009;29:6771–6779.[PMC free article] [PubMed] [Google Scholar]
48. Holtzman DM, Bales KR, Tenkova T, Fagan AM, Parsadanian M, Sartorius LJ, Mackey B, Olney J, McKeel D, Wozniak D, Paul SM. Apolipoprotein E isoform-dependent amyloid deposition and neuritic degeneration in a mouse model of Alzheimer’s disease. Proc Natl Acad Sci U S A. 2000;97:2892–2897.[PMC free article] [PubMed] [Google Scholar]
49. Deane R, Sagare A, Hamm K, Parisi M, Lane S, Finn MB, Holtzman DM, Zlokovic BV. apoE isoform-specific disruption of amyloid beta peptide clearance from mouse brain. J Clin Invest. 2008;118:4002–4013.[PMC free article] [PubMed] [Google Scholar]
50. Knudsen KA, Rosand J, Karluk D, Greenberg SM. Clinical diagnosis of cerebral amyloid angiopathy: Validation of the Boston criteria. Neurology. 2001;56:537–539. [PubMed] [Google Scholar]
Published online 2019 Jan 29. doi: 10.1002/advs.201802163
PMID: 30937277
Associated Data
ADVS-6-1802163-s001.pdf (1.5M)
Abstract
Methoxy‐functionalized triphenylamine‐imidazole derivatives that can simultaneously work as hole transport materials (HTMs) and interface‐modifiers are designed for high‐performance and stable perovskite solar cells (PSCs). Satisfying the fundamental electrical and optical properties as HTMs of p‐i‐n planar PSCs, their energy levels can be further tuned by the number of methoxy units for better alignment with those of perovskite, leading to efficient hole extraction. Moreover, when they are introduced between perovskite photoabsorber and low‐temperature solution‐processed NiOx interlayer, widely featured as an inorganic HTM but known to be vulnerable to interfacial defect generation and poor contact formation with perovskite, nitrogen and oxygen atoms in those organic molecules are found to work as Lewis bases that can passivate undercoordinated ion‐induced defects in the perovskite and NiOx layers inducing carrier recombination, and the improved interfaces are also beneficial to enhance the crystallinity of perovskite. The formation of Lewis adducts is directly observed by IR, Raman, and X‐ray photoelectron spectroscopy, and improved charge extraction and reduced recombination kinetics are confirmed by time‐resolved photoluminescence and transient photovoltage experiments. Moreover, UV‐blocking ability of the organic HTMs, the ameliorated interfacial property, and the improved crystallinity of perovskite significantly enhance the stability of PSCs under constant UV illumination in air without encapsulation.
Keywords: defect passivation, hole transport materials, perovskite solar cells, UV durability
1. Introduction
Organic–inorganic hybrid perovskite solar cell (PSC) has been receiving great attention in optoelectronic fields due to its fascinating optical and electrical features such as high absorption coefficient, long diffusion length and lifetime, and bipolar carrier transport.1, 2 Moreover, its cost‐effective low‐temperature solution processability makes it a good candidate for next‐generation solar cell.3 Along with the development of fabrication techniques, the advances in perovskite and charge transport materials including interfacial layers, the maximum power conversion efficiencies (PCEs) of PSCs have exceeded 22% now since their first application to solar cells.4, 5
The PSC device architectures can be classified into two types, a conventional n‐i‐p structure (planar and mesoscopic) and an inverted p‐i‐n structure.6, 7, 8 The record efficiency was from the n‐i‐p PSCs employing titanium oxide (TiO2) mesoporous (mp) structure as an electron‐transport layer (ETL), but the inevitable high‐temperature (over 400 °C) sintering process to create a conductive phase has hindered its flexible and large‐scale applications for future commercialization.4, 5, 9, 10 Furthermore, the planar n‐i‐p type PSCs without mp structure usually suffer from the severe hysteretic current density (J)–voltage (V) characteristics depending on the scan direction and rate.11, 12, 13 Those hysteretic behaviors are shown to be repressed in the p‐i‐n structure PSCs, because a little shorter hole diffusion length than the electron can be compensated in this architecture, improving hole extraction,14, 15 and fullerene derivative, often utilized as an ETL, is known to passivate defect sites of perovskite.16 In addition to this merit, the low‐temperature process feasibility has aroused extensive researches on this device architecture for low‐cost flexible PSCs.17
Employing suitable hole‐transporting layers (HTLs), which can reduce the energy losses of photogenerated carriers and minimize the interfacial charge recombination, is a prerequisite to achieve high‐performance p‐i‐n PSCs. Moreover, optimal selection of HTLs can enhance the device stability against constant full‐spectrum solar irradiation in ambient air by restraining the ion migration18, 19, 20 and improving the crystallinity of the active layer.21, 22, 23 The conventional organic HTLs such as poly(3,4‐ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS),24, 25 spiro‐OMeTAD,26, 27, 28 and poly(triarylamine) (PTAA)29, 30 may be able to impede the commercialization of PSCs due to their high‐price and dopant‐induced poor stability. By comparison, the inorganic HTLs such as CuGaO2,31 NiOx,32, 33, 34, 35 CuCrO2,36, 37 CuOx,38, 39 CuSbS2,40, 41, 42 and CuSCN43 exhibit low cost, superior stability and high carrier mobility without corrosion to the photoactive layer. Especially, low‐temperature‐processed NiOx is one of the most investigated due to its large bandgap with high transmittance in visible range, deep valence band edge, and intrinsic p‐type conductivity.32, 34 However, the low‐temperature solution processes usually increase the defects by deteriorating its crystallinity, which results in a series of deficiencies like reduced hole extraction and increased carrier recombination.44, 45 The interstitial vacancies and defects in NiOx thin films incur hole accumulation near the perovskite interface with the trap‐assisted nonradiative recombination losses, consequently decreasing the device performances. In addition, those defects induce rough contact with the perovskite layer, which degrades the crystallinity and morphology of perovskite layer with lower light absorption.14, 15, 46 Therefore, controlling the interface between NiOx HTL and perovskite photoactive layer can be an efficient way to improve the overall performance and stability of the PSCs having the low‐temperature‐processed NiOx.14, 47
In this work, a series of triphenylamine‐imidazole derivatives with different number of methoxy group, which can work as HTLs and surface modifiers simultaneously, are designed and applied to the interface between NiOx and perovskite in p‐i‐n type PSCs. From their electrical and optical properties, it is confirmed that those molecules are proper to the HTL in p‐i‐n PSCs (e.g., high hole mobility, proper energy level, and visible transparency) and the addition of methoxy moieties to backbone deepens their highest occupied molecular orbital (HOMO) levels that can improve built‐in potential of devices for better hole extraction. Moreover, N and O atoms in the HTL molecules are shown to form Lewis adducts with undercoordinated Ni and Pb ions in NiOx and perovskite, bifacially passivating those defect sites and improving the crystallinity of perovskite. Consequently, the PSCs having those organic HTLs even without dopant represent improved hole extraction and reduced charge recombination behaviors, and these enhanced properties are more discernible by increasing the number of methoxy units in the triphenylamine‐imidazole conjugation scaffold. Moreover, the UV‐blocking ability of the organic HTLs, the ameliorated interfacial property, and the improved crystalinity of perovskite significantly enhance the stability of PSCs under constant UV illumination in air without encapsulation.
2. Results and Discussion
2.1. Material Design
Figure1a shows the synthesized organic molecules sharing the same conjugated backbone with the different number of methoxy unit (TPI, TPI‐2MEO, TPI‐4MEO, and TPI‐6MEO having 0, 2, 4, and 6 methoxy units, respectively). The molecules could be synthesized by the straightforward two reaction steps from commercially available simple reagents with high reaction yield as explained in Figure S1 and Materials Synthesis section (Supporting Information). Materials characterization data such as 1H NMR (Figures S2, S4, S7, S10, and S13, Supporting Information), 13C NMR (Figures S3, S5, S8, S11, and S14, Supporting Information), and electrospray ionization‐mass spectrometry (ESI‐MS) (Figures S6, S9, S12, and S15, Supporting Information) are given in the Supporting Information. Particularly, their backbones were designed to be acceptor–donor–acceptor (A‐D‐A) type conjugated molecules. The electron‐rich triphenylamine dimer structure was utilized as the core of the materials due to its efficient hole transport mobility, and ambipolar imidazole derivatives were added as the acceptor due to their relative electron deficiency. In this design, the strong intermolecular interaction between backbones increases the probability of close molecule–molecule contacts facilitating charge carrier transport, and moreover, their optical and electrical properties can be easily tunable to be a HTM of PSCs.
a) Detailed chemical structures of the designed methoxy‐functionalized triphenylamine‐imidazole derivatives for dopant‐free hole‐transporting materials. b) Energy level of each layer of the perovskite solar cell. c) Measured logJ–logV plots of hole‐only devices under dark condition for the estimation of hole mobility of HTM. The calculated hole mobility values of TPI, TPI‐2MEO, TPI‐4MEO, and TPI‐6MEO from the SCLC method are 2.1 × 10−5, 2.5 × 10−5, 3.3 × 10−5, and 4.8 × 10−5 cm2 V−1 s−1, respectively.
Figure Figure1b1b represents that HOMO energy levels of designed molecules are well‐aligned with the valence band edge of perovskite (−5.4 eV) for efficient hole extraction and their lowest unoccupied molecular orbital (LUMO) energy levels are shallow enough to block the electron transfer from the perovskite. The HOMO energy levels of molecules were obtained by cyclic voltammetry (CV) results in Figure S16 (Supporting Information), and the LUMO energy levels were estimated by their absorption onset in Figure S17 (Supporting Information) and CV results. By introducing electron‐donating methoxy groups to the phenylimidazole fragments, which are electron‐accepting part in dumbbell‐shaped A‐D‐A molecules, it is confirmed that further fine‐tuning of the HOMO level is possible. The addition of methoxy groups to backbone deepens their HOMO levels (Figure (Figure1b),1b), advantageous to improve built‐in potential of devices enhancing their hole extraction property.7 Along with the variation of HOMO levels, methoxy groups are expected to function as Lewis base that can passivate defect sites of perovskite at the interface,48, 49 and this will be discussed with Raman, X‐ray photoelectron spectroscopy (XPS), and transient experiment results later again. Moreover, different from the conventional D‐A structured HTMs, applied to n‐i‐p PSCs having the absorption in the visible region, the bandgaps of molecules were adjusted for absorbing the light below 420 nm wavelength not to decrease the amount of incident photon transmitted into the perovskite photoabsorber in p‐i‐n configuration PSCs, and only the absorption around 280 nm wavelength region increased with the addition of methoxy unit, as shown in the absorbance spectra (Figure S17, Supporting Information). Meanwhile, the hole mobility values of the organic HTL materials were estimated by a space‐charge limited current method using the Mott–Gurney law: J = 9εoεrµV2/8L3, where εoεr is the permittivity of the component, µ is the carrier mobility, and L is the thickness,50 as shown in Figure Figure1c,1c, and their values were comparable to that of the conventional organic HTL material such as spiro‐OMeTAD in the literature.51
2.2. PSC Device Performances
Figure2a,b and Table1 show J–V characteristics of p‐i‐n PSC devices with and without organic HTL having the following configuration: ITO/NiOx/organic HTL/perovskite/PCBM/bathocuproine (BCP)/Ag. The scanning electron microscopy (SEM) images of the cross section of device and the surface of perovskite prepared on an organic HTL are shown in Figure Figure2c,d.2c,d. All devices with the organic HTL have improved performances (TPI: 15.92, TPI‐2MEO: 16.56, TPI‐4MEO: 17.59, and TPI‐6MEO: 18.42% PCE), compared to that without the organic HTL (14.57% PCE), and the PSCs with the organic HTL having the highest number of methoxy unit (TPI‐6MEO) represent the best performances, which are mainly ascribed to the enhanced short‐circuit current density (Jsc) and fill factor (FF). Especially, the slightly hysteretic J–V characteristics, observed from the pristine NiOx‐based PSCs, are suppressed in the PSCs with the organic HTLs, and J–V curves of a pristine NiOx‐based PSC and a TPI‐6MEO‐added PSC depending on the scan direction are in Figure Figure2b.2b. The stable operation of PSCs with the organic HTLs was further confirmed by the steady‐state output power at the maximum power point (MPP), and, for TPI‐6MEO‐added PSC as a representative, ≈18.3% of PCE was measured from MPP, comparable to that from its J–V characteristics (Figure S18, Supporting Information). Meanwhile, the external quantum efficiency (EQE) signals of PSCs show similar trend to their J–V characteristics as shown in Figure S19 (Supporting Information).
a) J–V curves of PSCs without and with organic HTLs (TPI, TPI‐2MEO, TPI‐4MEO, and TPI‐6MEO) on NiOx layer. Device configuration is ITO/NiOx/organic HTL/CH3NH3PbI3/PCBM/BCP/Ag. All data were measured at AM 1.5 G (100 mW cm−2 intensity). b) J–V curves of PSC built on TPI‐6MEO‐cast NiOx layer, scanned in forward and reverse directions. Inset figure is J–V curves of PSC built on bare NiOx layer, scanned in forward and reverse directions. c) The cross‐sectional SEM image of device (TPI‐6MEO‐applied PSC). d) The top‐view SEM image of perovskite prepared on organic HTL (TPI‐6MEO).
Table 1
Parameters of the PSCs with various HTMs measured by J–V characteristics
HTLa) | Jsca) [mA cm−2] | Voca) [V] | FFa) | PCEa) [%] |
---|---|---|---|---|
Pristine NiOx | 20.52 (21.31) | 0.99 (0.99) | 0.67 (0.69) | 13.61 (14.57) |
TPI | 20.66 (21.34) | 0.98 (0.99) | 0.74 (0.75) | 14.98 (15.92) |
TPI‐2MEO | 21.40 (22.01) | 0.98 (0.99) | 0.74 (0.76) | 15.52 (16.56) |
TPI‐4MEO | 22.58 (22.78) | 0.99 (0.99) | 0.76 (0.78) | 16.99 (17.59) |
TPI‐6MEO | 23.18 (23.31) | 0.97 (0.98) | 0.79 (0.81) | 17.76 (18.42) |
a)Numbers are average values of forward and reverse scan data, measured from at least over 30 devices for each condition (values in parentheses are from the best performing devices).
2.3. Defect Passivation of Organic HTM
To clarify the improved performances of PSCs with organic HTLs, the interface between NiOx and organic HTL was investigated by IR spectroscopy. The oxygen vacancy‐induced Ni3+ has been reported to be generated from the low‐temperature solution process,32, 52 and those defect sites could trap carriers, resulting in anomalous hysteresis of PSCs as well as degradation of their performances. IR spectra in Figure3a show that Ni—N vibration band at ≈526 cm−153 is observed from the organic HTL(TPI‐6MEO)‐casted NiOx film, which is not found in pristine NiOx and organic HTL‐only film. This suggests that N atoms in organic HTLs interact with NiOx, forming Ni—N bonds that can passivate the oxygen vacancy‐induced defects in NiOx.15
a) IR spectra of TPI‐6MEO‐only layer, NiOx‐only layer, and TPI‐6MEO‐cast NiOx layer on Si wafer substrate. b) Raman spectra of perovskite‐only layer and perovskite on TPI‐6MEO layer. c) XPS spectra showing the Pb 4f region of the perovskite on TPI‐6MEO film at 0.2 and 1.55 min sputtering time. d) XPS depth profiles of the perovskite on TPI‐6MEO film.
The methoxy functional groups in those organic HTL molecules are also expected to act as Lewis bases for defect‐healing by forming Lewis adducts with undercoordinated ions. From Raman spectroscopy results of perovskite‐cast organic HTL (TPI‐6MEO) film in Figure Figure3b,3b, we can clearly find peaks at 80 and 144 cm−1, related to Pb—O stretching,54 proving the formation of Lewis adducts between O atoms from electron‐donating methoxy groups of organic HTL and undercoordinated Pb ions of perovskite. To exclude the oxidation effect of perovskite that could induce Pb—O signals even without the organic HTL, the samples were characterized immediately after preparation within 1 h, and we observed that the pristine perovskite only started to generate Pb—O related signal (144 cm−1) after 4 h exposure to air (Figure S20, Supporting Information), which meant that Pb—O signals from the perovskite‐cast organic HTL were from the formation of Lewis adducts not from the oxidation of perovskite.
The formation of Lewis adduct was further confirmed by investigating the chemical states of Pb atoms at the interface between perovskite and organic HTL (TPI‐6MEO), obtained by XPS depth analysis of the perovskite‐cast organic HTL film. Figure Figure3d3d shows the variation of Pb 4f (from perovskite) and O 1s signal (from TPI‐6MEO) depending on the sputtering time, and the interfacial region of those two layers can be clarified by the abrupt decrease of Pb 4f signal with the increase of O 1s signal around 1.5–2.5 min of sputtering time. We especially focused on the binding energy shift of Pb 4f signals of the bulk perovskite (0.2 min sputtering time) to lower energy state at the interface (1.55 min sputtering time) (Figure (Figure3c),3c), and this represented the interaction of undercoordinated Pb ions in perovskite with Lewis bases in methoxy‐functionalized organic HTL at the interface,55 passivating possible halide vacancy defects in the perovskite.
The improved interfacial property between NiOx and perovskite with the organic HTLs is expected to provide better photocarrier extraction characteristics. Figure4a shows that the photoluminescence (PL) intensity of the perovskite on TPI‐6MEO‐cast NiOx is dramatically quenched, compared to that on bare NiOx, suggesting that interface passivation with the organic HTL is advantageous to efficiently transfer photocarriers from the perovskite to the NiOx layer. This feature was further confirmed by the decreased PL lifetimes of perovskite layers with the organic HTLs on NiOx (Figure (Figure4b),4b), which were calculated by convoluting decay curves using exponential functions (Table S1, Supporting Information). Figure Figure4b4b and Table S1 (Supporting Information) show that an averaged PL lifetime (τave) of perovskite on NiOx (τave = 18.1 ns) decreases with the organic HTLs, and it is getting shorter as the number of methoxy group of organic HTL that can passivate defects increases. The gradual deepening of HOMO energy levels of organic HTLs with the addition of methoxy units (Figure (Figure1b),1b), explained in former section, is also expected to be beneficial to improve the hole extraction characteristics. Consequently, perovskite on TPI‐6MEO‐cast NiOx, which has the highest number of electron‐donating methoxy group, has the shortest averaged PL lifetime of 6.1 ns, showing about threefold increase in the charge extraction rate, compared to that on bare NiOx. This trend is well‐matched with Jsc variation discussed earlier.
a) Steady‐state PL spectra of perovskite on NiOx layer and perovskite on TPI‐6MEO‐cast NOx layer. b) PL decay curves of perovskite on various organic HTM‐cast NiOx layer (quartz substrate, peak emission at 770 nm wavelength, and excitation at 670 nm wavelength). c) XRD patterns of perovskite on NiOx and perovskite on TPI‐6MEO‐cast NiOx. d) Absorption spectra of perovskite on NiOx and perovskite on TPI‐6MEO‐cast NiOx. e) Recombination lifetime versus light intensity plots of complete cells having various HTLs, calculated by TPV experiments.
The bifacial defect passivation of organic HTLs is also advantageous to improve the quality of perovskite layer. Figure Figure4c4c shows X‐ray diffraction (XRD) patterns of perovskite layers on pristine NiOx and organic HTL(TPI‐6MEO)‐cast NiOx, and the strong peaks at 14.5°, 28.4°, and 31.8°, which correspond to the (110), (220), and (310) planes of perovskite crystallites, are observed from both samples. Especially, the intensities of those peaks in organic HTLs‐applied perovskite are much higher than those in perovskite on bare NiOx, indicating better perovskite crystallinity with TPI‐6MEO modification. Moreover, the improved crystallinity of perovskite with organic HTL is beneficial to enhance the photon absorption as shown in Figure Figure44d.
The variation of charge recombination kinetics in the PSC devices with the organic HTLs was further investigated by transient photovoltage (TPV) measurement. TPV results in Figure Figure4e4e show the averaged charge recombination lifetimes of the PSCs with and without the organic HTLs, measured with an increment in Voc (from 0 to 1.0 Sun illumination). The recombination lifetimes of all the samples decrease with the increase of light illumination intensity, because the increased carrier concentration under higher light intensity accelerates the recombination process. TPV measurement is performed at open‐circuit condition, in which charge transport effects are minimized, and therefore transient signal can be regarded as being governed by recombination only, allowing the estimation of the carrier recombination.56 The PSC with TPI‐6MEO has the longest recombination lifetimes in the entire range of light intensity (i.e., Voc), and the lifetimes of PSCs are prolonged as the number of methoxy unit in organic HTL increase, which indicates the retardation of charge recombination by the defect passivation and better energy level alignment with methoxy units.
2.4. Device Stability
In addition to high efficiency, long‐term device stability is another crucial parameter, as it determines its suitability for commercialization. We traced the PCEs of the PSCs with and without TPI‐6MEO for 300 h of storage in N2 box without encapsulation. As shown in Figure5a, the PSC without organic HTL maintained about 50% of its original PCE, whereas the TPI‐6MEO‐applied PSC retained about 85% of its initial PCE within the same period. It is expected that the enhanced crystallinity of the perovskite layer with the ameliorated interfacial contact, preventing the permeation of oxygen and water into the perovskite film, improves the stability of TPI‐6MEO‐applied PSC.
a) Stability of PSCs having NiOx‐only and TPI‐6MEO‐cast NiOx as the HTL stored in N2 box without encapsulation. b) Photostability test results of PSCs having NiOx‐only and TPI‐6MEO‐cast NiOx as the HTL under constant UV illumination (365 nm, 500 mW cm−2) without encapsulation in air. XRD pattern variation of c) perovskite on NiOx and d) perovskite on TPI‐6MEO‐cast NiOx according to continuous UV exposure time (365 nm, 500 mW cm−2) without encapsulation in air.
Perovskite materials are known to be easily decomposed under sunlight, especially UV light. Moreover, UV irradiation accelerates the degradation of the metal oxide–perovskite interface, increasing defect sites for charge recombination. The devices with and without the TPI‐6MEO layer were exposed to continuous UV light (365 nm, 500 mW cm−2) in the ambient air without encapsulation, and their performances were measured every 3 h to verify the effect of the organic HTL on the UV stability of PSCs. Figure Figure5b5b shows that the PSC without organic HTL is degraded dramatically after UV light exposure for 21 h, retaining only 30% of its initial performance, but the PSC with TPI‐6MEO interlayer preserves ≈85% of its original PCE. This enhanced UV durability with TPI‐6MEO was further confirmed by XRD signal variation of perovskite films with and without TPI‐6MEO under UV irradiation. Figure Figure5c,d5c,d presents the XRD spectra of the perovskite films before and after UV irradiation for 6 and 12 h. Before UV light illumination, the XRD patterns of both two samples are almost identical, where the main diffraction peaks at 14.5°, 28.4°, and 31.8° are attributed to the (110), (220), and (310) planes of perovskite films. However, after 6 h UV aging, a diffraction peak centered at 12.49°, which is ascribed to the (001) diffraction peak of PbI2, is newly observed in the XRD patterns of perovskite film without TPI‐6MEO, and the intensity of this peak increases after 12 h UV illumination, indicating the decomposition of perovskite material into MAI and PbI2 under UV light. In contrast, the XRD patterns of perovskite film with TPI‐6MEO under the same UV illumination condition retain the original diffraction pattern after 6 h and show only small peak centered at 12.49° even after 12 h UV irradiation. The enhanced UV stability of PSC with TPI‐6MEO is attributed to the strong absorption of TPI‐6MEO under 400 nm wavelength light, blocking the UV light into perovskite layer, the passivated defect sites at the interface, and the improved perovskite crystallinity.
3. Conclusion
The defect management of perovskite and its interface with the interlayer is one of the crucial factors to achieve high‐performance and high‐stability PSCs. In this work, organic HTMs, which not only have proper optical and electrical properties for p‐i‐n PSCs without dopants but also bifacially passivate defect sites at the interfaces of perovskite and NiOx layers by forming Lewis adducts, are demonstrated. Those HTLs are also advantageous to enhance the crystallinity of following perovskite photoabsorber. Consequently, we can observe the improved charge extraction and suppressed recombination properties from the organic HTL‐applied PSCs, providing superior performances. Moreover, the UV‐blocking ability of the organic HTLs, the ameliorated interfacial property, and the improved crystallinity of perovskite significantly enhance the stability of PSCs under constant UV illumination in air without encapsulation.
4. Experimental Section
Materials and Characterization: N,N,N′,N′‐tetraphenylbenzidine, phosphorus oxychloride (POCl3), N,N′‐dimethylformamide (DMF) anhydrous, aniline, 4‐methoxyaniline, benzil, 4,4′‐dimethoxybenzil, ammonium acetate, and glacial acetic acid were purchased from Sigma Aldrich and used without further purification. All glasswares, syringes, magnetic stirrer bars, and needles were thoroughly dried before use. Reactions were monitored using thin layer chromatography (TLC) with TLC plates (silica gel 60 F254, Merck Co.). Silica gel column chromatography was performed with silica gel 60 (particle size 0.0063–0.200 mm, Merck). 1H NMR and 13C NMR spectra were obtained from Bruker (Fourier 300 MHz) and Oxford (Fourier 400 MHz) NMR spectrometers, and the chemical shifts (δ) and coupling constant (J) were expressed in ppm and Hertz, respectively. ESI‐MS spectra were obtained on a UHPLC/tandem mass spectrometer (1290 infinity II/Qtrap 6500). Absorption spectra were monitored by UV‐2600 UV–vis spectrophotometer (SHIMADZU) in the wavelength range 200–900 nm. Cyclic voltammetry analysis was performed on a potentiostat/galvanostat model 273A (Princeton Applied Research) for obtaining the potential. A supporting electrolyte was made to 0.1 m tetra‐n‐butylammonium tetrafluoroborate solution in dichloromethane. Parameters were set to scan rate of 100 mV s−1 and vertex potential of 2.5 V.
Device Fabrication: PSC devices were fabricated on patterned indium‐doped tin oxide (ITO) coated glass. ITO was etched with 35% HCl and zinc powder. These substrates were cleaned using ultrasonication in acetone, isopropyl alcohol (IPA), and deionized water for about 3 min, respectively. Substrates were then treated by oxygen plasma for the enhanced wettability of the following HTM solution. NiOx nanoparticle solution (20 mg mL−1) was spun on the ITO layer (2000 rpm for 20 s) and annealed at 100 °C for 10 min. NiOx nanoparticle synthesis procedures are described elsewhere.57 Triphenylamine‐imidazole‐based HTMs were dissolved in chloroform at a concentration of 20 mg mL−1, and then spin‐cast on the NiOx layer at 2000 rpm for 20 s. CH3NH3PbI3 perovskite solution was prepared by dissolving PbI2 (Sigma Aldrich) and CH3NH3I (1:1 molar ratio) in the solvent mixture of DMF and dimethylsulfoxide (DMSO) (9:1 v/v) for a total concentration of 1.6 m in a N2 atmosphere. The solution was stirred at room temperature for at least 2 h before being used, then filtered via polytetrafluoroethlyene (PTFE) filter (0.45 µm). The perovskite layer was formed onto the HTM by a one‐step spin‐casting process at 4000 rpm for 25 s. After 9 s, the substrate was treated with ether (0.5 mL) by drop‐casting. The substrate was annealed on a hot plate at 100 °C for 10 min. PCBM (20 mg mL−1 in chlorobenzene) was spin‐cast on the perovskite layer. BCP (0.5 mg mL−1 in ethanol) solution was prepared and spun at 5000 rpm for 60 s, and then dried 10 min. On the last stage, samples were transferred into a thermal evaporator and Ag (100 nm) was deposited at a pressure of 5 × 10−6 torr giving a following device configuration: ITO/NiOx/organic HTL/CH3NH3PbI3/PCBM/BCP/Ag. The hole‐only devices to estimate the hole mobility values of organic HTLs were fabricated by utilizing MoO3 having high work function to block the injection of electrons from the Ag electrode (ITO/NiOx/organic HTL/MoO3/Ag).
Device Characterization: A solar simulator (PEC‐L01, Peccell Technologies, Inc.) with AM 1.5 G illumination provides 100 mW cm−2 of illumination on the PV cells. The intensity was calibrated using a NREL‐certified Si photodiode, equipped with an infrared cutoff filter (KG5) to reduce spectral mismatch. J–V characteristics were obtained using an Ivium Technology Ivium CompactStat by scanning at a 0.05 V s−1 scan rate. The EQE was measured at short‐circuit condition using an ABET Technology 10500 solar simulator as the light source and a SPECTRO Mmac‐200 as the light solution.
FTIR, Raman, XRD, and XPS Characterization: FTIR spectra were recorded by a Nicolet 5700 instrument (Thermofisher Scientific, USA) from 4000 to 400 cm−1 with a resolution of 4 cm−1. Samples have the following configurations: Si wafer/NiOx, Si wafer/TPI‐6MEO, and Si wafer/NiOx/TPI‐6MEO. Raman spectra were obtained by a high‐resolution Raman spectrometer (LabRam HR Evolution, HORIBA, Japan) with a pumped laser of 532 nm and the resolution of 600 gr mm−1. Raman spectra were measured for 20 s by a 100× objective, resulting in a laser spot diameter less than 1 µm on the sample. Raman samples have the following configurations: glass/TPI‐6MEO/perovskite and glass/perovskite. XRD patterns were collected via an Ultima III (Rigaku) diffractometer using copper Kα radiation. X‐ray pattern was measured by step scanning at angular intervals of 0.08° from 10° to 90°. 2θ scans were obtained from samples having the following configurations: glass/NiOx/perovskite and glass/NiOx/TPI‐6MEO/perovskite). XPS measurements were carried out using a PHI 5000 VersaProbe spectrometer equipped with an Al Kα X‐ray source (1486.6 eV). An Ar+ ion gun with a 2 kV beam voltage in a 2 × 2 mm raster area (yielding an equivalent sputtering rate of 8 nm min−1 of SiO2) was used to XPS depth profiling. All binding energies in the XPS data were calibrated with reference to the C−C bond in the C 1s. The sample for XPS measurement has a glass/TPI‐6MEO/perovskite configuration.
Time‐Resolved Photoluminescence (TRPL) and TPV: TRPL curves were recorded using a commercial TCSPC system (FluoTime 200, PicoQuant). Samples were excited by using a picosecond diode laser of 670 nm (LDH‐P‐C‐670, PicoQuant) with a repetition rate of 4 MHz. The emitted PL signal was accumulated via a fast photomultiplier tube (PMT) detector (PMA 182, PicoQuant) with a magic angle (54.7°) arrangement. The incident angle of excitation pulse was set to be about 30° with respect to the sample. The resulting instrumental response function was 160 ps in full width at half maximum. The PL decays were measured at the emission peak (770 nm) for perovskite. In addition, a cutoff filter (FF01‐692 nm, Semrock) was applied to block the scattering. The transient photovoltage measurement was conducted by a nanosecond OPO laser system (INDI‐40‐10, Spectra‐Physics) with Nd:YAG laser and a background illumination from Xe lamp (LS‐150‐XE ABET). An attenuated laser pulse at 550 nm laser pulse with a pulse width of 120 fs was used as a small perturbation to the background illumination for generating an additional amount of charge on the devices. The laser‐pulse‐induced photovoltage variation was smaller than 3% of the Voc not exceeding 20 mV produced by the background illumination. The device was connected to a digital oscilloscope (DSO‐X 3054A, Agilent) with BNC cables, and the input impedance of the oscilloscope was set to 1 MΩ to form the open‐circuit conditions. The bias light intensity was adjusted by neutral density filters (NDC‐100C‐4M) for various Voc values. The initial light intensity from the Xe lamp was modified using power meter to be equivalent to 1 Sun (100 mA cm−2).
Conflict of Interest
The authors declare no conflict of interest.
Supporting information
Supplementary
Acknowledgements
Z.L., B.H.J., and S.J.H. contributed equally to this work. This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF‐2017R1D1A1B03034711 and NRF‐2017R1D1A3B03033045). This work was also partially supported by the Korea Institute of Energy Technology Evaluation and Planning (KETEP) and granted financial resource from the MOTIE (20164030201380).
Notes
Li Z., Jo B. H., Hwang S. J., Kim T. H., Somasundaram S., Kamaraj E., Bang J., Ahn T. K., Park S., Park H. J., Adv. Sci.2019, 6, 180216310.1002/advs.201802163 [CrossRef] [Google Scholar]
Contributor Information
Sanghyuk Park, Email: rk.ca.ujgnok@0290kraps.
Hui Joon Park, Email: rk.ca.uoja@noojiuh.
References
1. Stranks S. D., Eperon G. E., Grancini G. E., Menelaou C., Alcocer M. J. P., Leijtens T., Herz L. M., Petrozza A., Snaith H. J., Science2013, 342, 341. [PubMed] [Google Scholar]
2. Xing G., Mathews N., Sun S., Lim S. S., Lam Y. M., Gratzel M., Mhaisalkar S., Sum T. C., Science2013, 342, 344. [PubMed] [Google Scholar]
3. Green M. A., Hishikawa Y., Dunlop E. D., Levi D. H., Hohl‐Ebinger J., Ho‐Baillie A. W. Y., Prog. Photovoltaics2018, 26, 3. [Google Scholar]
4. Yang W. S., Park B. W., Jung E. H., Jeon N. J., Kim Y. C., Lee D. U., Shin S. S., Seo J., Kim E. K., Noh J. H., Seok S. I., Science2017, 356, 1376. [PubMed] [Google Scholar]
5. Jeon N. J., Na H., Jung E. H., Yang T. Y., Lee Y. G., Kim G., Shin H. W., Seok S. I., Lee J., Seo J., Nat. Energy2018, 3, 682. [Google Scholar]
6. Liu M., Johnston M. B., Snaith H. J., Nature2013, 501, 395. [PubMed] [Google Scholar]
7. Ok S. A., Jo B., Somasundaram S., Woo H. J., Lee D. W., Li Z., Kim B. G., Kim J. H., Song Y. J., Ahn T. K., Park S., Park H. J., Nat. Commun.2018, 9, 4537. [PMC free article] [PubMed] [Google Scholar]
8. Luo D., Yang W., Wang Z., Sadhanala A., Hu Q., Su R., Shivanna R., Trindade G. F., Watts J. F., Xu Z., Liu T., Chen K., Ye F., Wu P., Zhao L., Wu J., Tu Y., Zhang Y., Yang X., Zhang W., Friend R. H., Gong Q., Snaith H. J., Zhu R., Science2018, 360, 1442. [PubMed] [Google Scholar]
9. Saliba M., Matsui T., Domanski K., Seo J. Y., Ummadisingu A., Zakeeruddin S. M., Correa‐Baena J. P., Tress W. R., Abate A., Hagfeldt A., Gratzel M., Science2016, 354, 206. [PubMed] [Google Scholar]
10. Thakur U., Kwon U., Hasan M. M., Yin W., Kim D., Ha N. Y., Lee S., Ahn T. K., Park H. J., Sci. Rep.2016, 6, 35994. [PMC free article] [PubMed] [Google Scholar]
11. Unger E. L., Hoke E. T., Bailie C. D., Nguyen W. H., Bowring A. R., Heumüller T., Christoforo M. G., McGehee M. D., Energy Environ. Sci.2014, 7, 3690. [Google Scholar]
12. Snaith H. J., Abate A., Ball J. M., Eperon G. E., Leijtens T., Noel N. K., Stranks S. D., Wang J. T., Wojciechowski K., Zhang W., J. Phys. Chem. Lett.2014, 5, 1511. [PubMed] [Google Scholar]
13. Bergmann V. W., Weber S. A. L., Ramos F. J., Nazeeruddin M. K., Gratzel M., Li D., Domanski A. L., Lieberwirth I., Ahmad S., Berger R., Nat. Commun.2014, 5, 5001. [PubMed] [Google Scholar]
14. Wang Q., Chueh C. C., Zhao T., Cheng J., Eslamian M., Choy W. C. H., Jen A. K. Y., ChemSusChem2017, 10, 3794. [PubMed] [Google Scholar]
15. Bai Y., Chen H., Xiao S., Xue Q., Zhang T., Zhu Z., Li Q., Hu C., Yang Y., Hu Z., Huang F., Wong K. S., Yip H.‐L., Yang S., Adv. Funct. Mater.2016, 26, 2950. [Google Scholar]
16. Shao Y., Xiao Z., Bi C., Yoan Y., Huang J., Nat. Commun.2014, 5, 5784. [PubMed] [Google Scholar]
17. Wu Z., Li P., Zhang Y., Zheng Z., Small Methods2018, 2, 1800031. [Google Scholar]
18. Xu J., Buin A., Ip A. H., Li W., Voznyy O., Comin R., Yuan M., Jeon S., Ning Z., McDowell J. J., Kanjanaboos P., Sun J. P., Lan X., Quan L. N., Kim D. H., Hill I. G., Maksymovych P., Sargent E. H., Nat. Commun.2015, 6, 7081. [PMC free article] [PubMed] [Google Scholar]
19. Eames C., Frost J. M., Barnes P. R. F., O'Regan B. C., Walsh A., Islam M. S., Nat. Commun.2015, 6, 7497. [PMC free article] [PubMed] [Google Scholar]
20. Yuan Y., Huang J., Acc. Chem. Res.2016, 49, 286. [PubMed] [Google Scholar]
21. Mosconi E., Meggiolaro D., Snaith H. J., Stranks S. D., De Angelis F., Energy Environ. Sci.2016, 9, 3180. [Google Scholar]
22. Yu H., Lu H., Xie F., Zhou S., Zhao N., Adv. Funct. Mater.2016, 26, 1411. [Google Scholar]
23. Wang F., Geng W., Zhou Y., Fang H. H., Tong C. J., Loi M. A., Liu L. M., Zhao N., Adv. Mater.2016, 28, 9986. [PubMed] [Google Scholar]
24. Huang D., Goh T., Kong J., Zheng Y., Zhao S., Xu Z., Taylor A. D., Nanoscale2017, 9, 4236. [PubMed] [Google Scholar]
25. Liu D., Li Y., Yuan J., Hong Q., Shi G., Yuan D., Wei J., Huang C., Tang J., Fung M.‐K., J. Mater. Chem. A2017, 5, 5701. [Google Scholar]
26. Yang G., Chen C., Yao F., Chen Z., Zhang Q., Zheng X., Ma J., Lei H., Qin P., Xiong L., Ke W., Li G., Yan Y., Fang G., Adv. Mater.2018, 30, 1706023. [PubMed] [Google Scholar]
27. Anaraki E. H., Kermanpur A., Steier L., Domanski K., Matsui T., Tress W., Saliba M., Abate A., Grätzel M., Hagfeldt A., Correa‐Baena J.‐P., Energy Environ. Sci.2016, 9, 3128. [Google Scholar]
28. Jiang Q., Chu Z., Wang P., Yang X., Liu H., Wang Y., Yin Z., Wu J., Zhang X., You J., Adv. Mater.2017, 29, 1703852. [Google Scholar]
29. Zhang S., Stolterfoht M., Armin A., Lin Q., Zu F., Sobus J., Jin H., Koch N., Meredith P., Burn P. L., Neher D., ACS Appl. Mater. Interfaces2018, 10, 21681. [PubMed] [Google Scholar]
30. Wang Q., Bi C., Huang J., Nano Energy2015, 15, 275. [Google Scholar]
31. Papadas I. T., Savva A., Ioakeimidis A., Eleftheriou P., Armatas G. S., Choulis S. A., Mater. Today Energy2018, 8, 57. [Google Scholar]
32. Kwon U., Kim B. G., Nguyen D. C., Park J. H., Ha N. Y., Kim S. J., Ko S. H., Lee S., Lee D., Park H. J., Sci. Rep.2016, 6, 30759. [PMC free article] [PubMed] [Google Scholar]
33. Flynn C. J., McCullough S. M., Li L., Donley C. L., Kanai Y., Cahoon J. F., J. Phys. Chem. C2016, 120, 16568. [Google Scholar]
34. Cao J., Yu H., Zhou S., Qin M., Lau T.‐K., Lu X., Zhao N., Wong C.‐P., J. Mater. Chem. A2017, 5, 11071. [Google Scholar]
35. Seo S., Park I. J., Kim M., Lee S., Bae C., Jung H. S., Park N.‐G., Kim J. Y., Shin H., Nanoscale2016, 8, 11403. [PubMed] [Google Scholar]
36. Xiong D., Xu Z., Zeng X., Zhang W., Chen W., Xu X., Wang M., Cheng Y.‐B., J. Mater. Chem.2012, 22, 24760. [Google Scholar]
37. Zhang H., Wang H., Zhu H., Chueh C.‐C., Chen W., Yang S., Jen A. K.‐Y., Adv. Energy Mater.2018, 81702762. [Google Scholar]
38. Rao H., Ye S., Sun W., Yan W., Li Y., Peng H., Liu Z., Bian Z., Li Y., Huang C., Nano Energy2016, 27, 51. [Google Scholar]
39. Yu Z., Liu W., Fu W., Zhang Z., Yang W., Wang S., Li H., Xu M., Chen H., J. Mater. Chem. A2016, 4, 5130. [Google Scholar]
40. Pal M., Luna Y. T., González R. S., Mathews N. R., Paraguay‐Delgado F., Pal U., Mater. Des.2017, 136, 165. [Google Scholar]
41. Moosakhani S., Alvani A. A. S., Mohammadpour R., Ge Y., Hannula S.‐P., J. Alloys Compd.2018, 736, 190. [Google Scholar]
42. Moosakhani S., Alvani A. A. S., Mohammadpour R., Hannula P.‐M., Ge Y., Hannula S.‐P., Mater. Lett.2018, 215, 157. [Google Scholar]
43. Wijeyasinghe N., Regoutz A., Eisner F., Du T., Tsetseris L., Lin Y.‐H., Faber H., Pattanasattayavong P., Li J., Yan F., McLachlan M. A., Payne D. J., Heeney M., Anthopoulos T. D., Adv. Funct. Mater.2017, 27, 1701818. [Google Scholar]
44. Corani A., Li M.‐H., Shen P.‐S., Chen P., Guo T.‐F., El Nahhas A., Zheng K., Yartsev A., Sundstrom V., Ponseca C. S. Jr., J. Phys. Chem. Lett.2016, 7, 1096. [PubMed] [Google Scholar]
45. Fan R., Huang Y., Wang L., Li L., Zheng G., Zhou H., Adv. Energy Mater.2016, 6, 1600460. [Google Scholar]
46. Hou Y., Chen W., Baran D., Stubhan T., Luechinger N. A., Hartmeier B., Richter M., Min J., Chen S., Quiroz C. O. R., Li N., Zhang H., Heumueller T., Matt G. J., Osvet A., Forberich K., Zhang Z.‐G., Li Y., Winter B., Schweizer P., Spiecker E., Brabec C. J., Adv. Mater.2016, 28, 5112. [PubMed] [Google Scholar]
47. Nie W., Tsai H., Blancon J.‐C., Liu F., Stoumpos C. C., Traore B., Kepenekian M., Durand O., Katan C., Tretiak S., Crochet J., Ajayan P. M., Kanatzidis M. G., Even J., Mohite A. D., Adv. Mater.2018, 30, 1703879. [Google Scholar]
48. Noel N. K., Abate A., Stranks S. D., Parrott E. S., Burlakov V. M., Goriely A., Snaith H. J., ACS Nano2014, 8, 9815. [PubMed] [Google Scholar]
49. Park S. J., Jeon S., Lee I. K., Zhang J., Jeong H., Park J.‐Y., Bang J., Ahn T. K., Shin H.‐W., Kim B.‐G., Park H. J., J. Mater. Chem. A2017, 5, 13220. [Google Scholar]
50. Park H. J., Kim H., Lee J. Y., Lee T., Guo L. J., Energy Environ. Sci.2013, 6, 2203. [Google Scholar]
51. Bi D., Xu B., Gao P., Sun L., Gratzel M., Hagfeldt A., Nano Energy2016, 23, 138. [Google Scholar]
52. You J., Meng L., Song T.‐B., Guo T.‐F., Yang Y. M., Chang W.‐H., Hong Z., Chen H., Zhou H., Chen Q., Liu Y., De Marco N., Yang Y., Nat. Nanotechnol.2016, 11, 75. [PubMed] [Google Scholar]
53. Yang B., Li J., Zhang L., Xu G., Analyst2016, 141, 5822. [PubMed] [Google Scholar]
54. Qin P.‐L., Yang G., Ren Z.‐W., Cheung S. H., So S. K., Chen L., Hao J., Hou J., Li G., Adv. Mater.2018, 30, 1706126. [PubMed] [Google Scholar]
55. Zheng H., Zhu L., Hu L., Yang S., Chen S., Alsaedi A., Hayat T., Huang Y., Pan X., Dai S., J. Mater. Chem. A2018, 6, 9930. [Google Scholar]
56. MacKenzie R. C. I., Shuttle C. G., Chabinyc M. L., Nelson J., Adv. Energy Mater.2012, 2, 662. [Google Scholar]
57. Lee K.‐T., Jang J.‐Y., Park S. J., Ok S. A., Park H. J., Nanoscale2017, 9, 13983. [PubMed] [Google Scholar]
Articles from Advanced Science are provided here courtesy of Wiley-Blackwell