Construction and investigation of photo-switch property of azobenzene-bridged pillar[5]arene-based [3]rotaxanes

Article abstract of DOI:10.1016/j.cclet.2020.11.020

Series of azobenzene-bridged pillar[5]arene-based [3]rotaxanes with different alkyl chain length of guest molecules were constructed by threading-endcapping method with alkylenetriazole as axile and tetrahydrochromene as endcapping group. The encapsulation of pillar[5]arenes were proved by high-resolution mass, ¹H NMR and NOESY spectra. The photo-responsive property were examined by irradiation of the synthesized [3]rotaxanes with 365 nm and blue light LED, which caused trans to cis and cis to trans isomerization, respectively. Irradiation of corresponding model guest compounds without pillar[5]arene encapsulation resulted in near completely trans to cis and cis to trans isomerization, indicating the existence of pillar[5]arenes is the determining factor for the comprised photo isomerization efficiency.

Full text of DOI:10.1016/j.cclet.2020.11.020

Journal Pre-proof
Construction and investigation of photo-switch property of
azobenzene-bridged pillar[5]arene-based [3]rotaxanes
Hui Chong, Cuiyun Nie, Lihong Wang, Sicong Wang, Ying Han,
Yang Wang, Chengyin Wang, Chaoguo Yan
PII:
S1001-8417(20)30652-5
https://doi.org/10.1016/j.cclet.2020.11.020
CCLET 5960
DOI:
Reference:
To appear in:
Chinese Chemical Letters
Accepted Date:
10 November 2020
Please cite this article as: Chong H, Nie C, Wang L, Wang S, Han Y, Wang Y, Wang C, Yan C,
Construction and investigation of photo-switch property of azobenzene-bridged
pillar[5]arene-based [3]rotaxanes, Chinese Chemical Letters (2020),
doi: https://doi.org/10.1016/j.cclet.2020.11.020
This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
© 2020 Published by Elsevier.

Please donot adjust the margins
Communication
Construction and investigation of photo-switch property of azobenzene-bridged
pillar[5]arene-based [3]rotaxanes
Hui Chong, Cuiyun Nie, Lihong Wang, Sicong Wang, Ying Han* hanying@yzu.edu.cn, Yang Wang,
Chengyin Wang, Chaoguo Yancgyan@yzu.edu.cn
College of Chemistry & Chemical Engineering, Yangzhou University, Yangzhou 225002, China
 Corresponding authors.
Email addresses : (Y. Han), (C. Yan)
Graphical abstract
A series of unique azobenzene-bridged pillar[5]arene-based [3]rotaxanes with different alkyl chain length of guest were constructed by
threading-endcapping method with alkylenetriazole as axil and tetrahydrochromene as endcapping group.
ABSTRACT
Series of azobenzene-bridged pillar[5]arene-based [3]rotaxanes with different alkyl chain length of guest molecules were
constructed by threading-endcapping method with alkylenetriazole as axile and tetrahydrochromene as endcapping group. The
1
encapsulation of pillar[5]arenes were proved by high-resolution mass, H NMR and NOESY spectra. The photo-responsive property
were examined by irradiation of the synthesized [3]rotaxanes with 365 nm and blue light LED, which caused trans to cis and cis to
trans isomerization, respectively. Irradiation of corresponding model guest compounds without pillar[5]arene encapsulation resulted in
near completely trans to cis and cis to trans isomerization, indicating the existence of pillar[5]arenes is the determining factor for the
comprised photo isomerization efficiency
Keywords:Pillar[5]arene [3]Rotaxane Azobenzene Supramolecule Photo-switch
Stimuli-responsive supramolecular systems nowadays have gained intense research interests as they are potential building blocks for
fabrication of memory storage and drug delivery systems, chemical sensors and molecular machines [1-4]. In this aspect, the
mechanically interlock molecules (MIMs) are attractive entities because of their potential stimuli caused the shuttle behaviors, which
enable their applications as molecular device and machine [5]. A plethora of external stimuli including temperature, pH, light and
redox were commonly employed in the fabrication of such functional systems [6-12]. Among them, light is the most facile and
promising external stimuli and has attracted considerable attenations [13-15]. Azobenzene was recognized as one of the key light
responsive motifs among the photo active moieties because of light triggered trans/cis isomerization, which could render substantial
structure and property change of azobenzene [15]. Therefore, azobenzene derivates have been widely employed as guest of light-
responsive MIMs [16].
Traditional host molecules for fabricating rotaxanes included crown ethers, cyclodextrins, and cucurbit[n]urils [17]. The synthesis
and photo-responsive property of cucurbit[7]uril and azobenzene based [3]rotaxane has been reported. The systems exhibited
molecular shuttle property upon irradiation of UV/vis light [18]. The novel macrocyclic structure of pillar[n]arenes (n = 5, 6), first
reported in 2008 and described as “fascinating cyclophanes with a bright future”, have been proved to be the promising candidate host
for rotaxane and catananes [15]. Recently, some rotaxanes composed of azobenzene derivates and pillar[n]arenes hav been reported in
literatures [19-21]. The photo-switch properties together with the function exploration of pillar[n]arenes-based azobenzene-bearing
rotaxanes received considerable attentions [15]. Yet, current related investigations mainly focused on pillar[n]arene (n = 5, 6) based
azobenzene bridged [1] and [2]rotaxanes. The light-responsive property of bis-pillar[n]arene and azobenzene based [3]rotaxanes
remain much unexplored. Herein, we report the design, syntheses and photo-switch property of pillar[5]arene-based azobenzene-
bridged [3]rotaxanes.
We used a straight-forward strategy to synthesize the pillar[5]arene-based azobenzene-bridged [3]rotaxanes (Scheme 1). The
alkylation of 4,4'-dihoxylazobenzene with 3-bromoprop-1-yne readily afforded the triple bond bearing photo-active segments. The
triple bonds attached to the functional core facilitate subsequently constructing of precursors of guest molecules by Sharpless
———

Please donot adjust the margins
copper(I)-catalyzed Sharpless alkyne-azide click reaction (Scheme 1). The triazole groups formed in this step could help “fix” the
pillar[5]arenes by multiple hydrogen bonds interactions. With the consideration that alkyl chain length on the guest molecule may have
different influence on the photo-assisted shuttle property of the supramolecules, we synthesized three azide-end capped segments with
varying alkyl chain lengths (C4, C6 and C10). The symmetrical azobenzene bearing threads (5a-5c) was also synthesized as standard.
The alkyl chain length did not show obvious influence on the yield of the click reaction. The threading of pillar[5]arene onto the guest
molecule was achieved in chloroform solution. And the benzaldehyde end capped pseudo rotaxanes was immediately reacted with 5,5-
dimethylcyclohexane-1,3-dione and malononitrile to install the bulky stopper to block the encapsulated pillar[5]arenes. In contrast to
the situation of 6a and 6b, the yield for 6c was obviously reduced (24% for 6a, 26% for 6b and 8% for 6c), may suggest a relatively
low threading efficiency of pillar[5]arene in the case of molecular thread of C10 chain.
In addition to these [3]rotaxanes, we also synthesized bulky stopper installed guest molecules (5a-5c) without threading of
1
pillar[5]arenes in order to examine the influence of alkyl chain length on photo-responsive property in advance. H, 13C NMR and
high-resolution mass spectra data for these compounds were shown in supporting compound 6a exhibited obvious upfield δ
movements in information. As shown in Fig. 1, the protons Hd-Hg of contrast to 5a. The upfield δ value of Hd (Δδ = -0.447 for Hd)
suggested the formation of C-H···O hydrogen bonds between triazole segment and the oxygen atoms on the pillar[5]arene as
documented in the literature [22]. The upfield δ movements of protons He to Hh suggested inclusion of the methylene units within the
cavity of pillar[5]arene as the result of the encapsulation-induced shielding effects [23]. These results clearly indicated formation of
1
unique [3]rotaxanes. In addition to the H NMR spectra, the exact molecular weight of 6a measured by high-resolution mass matched
the calculated value, further indicating the formation of [3]rotaxanes (Figs. S22, S25 and S28 in Supporting information).
Scheme 1. Synthetic route for modes 5a-5c and [3]rotaxanes 6a-6c
Fig. 1. Partial 1H NMR spectra of 5a and 6a (CDCl3, 400 MHz).
It should be noticed that the small cavity size of pillar[5]arene (inner diameter of around 4.7Å) typically does not allow it to
encapsulate the azobenzene segment (the diameter of benzene ring is around 5.0 Å). Therefore, it is highly possible that the

Please donot adjust the margins
pillar[5]arene would be squeezed to locate on the alkyl chain segment of guest molecules. After the threading of pillar[5]arenes, δ
values of Hf and Hg on the alkyl chain 6a displayed dramatically upfield in movements among all the alkyl chain protons (Δδ = -3.370
and -3.180 for Hf and Hg; Δδ = -0.065 and -0.582 for He and Hh). Such alternation in NMR indicated strong C-H···π interactions
between the pillar[5]arenes [24]. On the other hand, the splitted aromatic protons on pillar[5]arene (δ = 6.99 and 6.86 for aromatic
1
protons in 6a suggested different chemical environments of the two rims of pillar[5]arenes upon encapsulation. Based on these H
NMR results, it was reasonable to predict the pillar[5]arenes located near the triazole segments of the guest molecule with the cavity
covering He-g of the alkyl chain. This configuration was further supported by the DFT optimized structures of the compound 5a and 6a,
based on which the computed upfielded δ for Hf and Hg is ~3 ppm, in good agreement with the experimental observation mentioned
above (Fig. 2).
Fig. 2. Optimized structures of 5a and 6a in dichloromethane solution (Hydrogens Hf and Hg are highlighted in cyan).
The δ values of alkyl chain protons of the guest molecules of 6b and 6c exhibited similar upshift pattern in contrast to the
corresponding model guest compounds. This also indicated the formation of corresponding [3]rotaxanes for the compounds 6b and 6c,
and was also supported by high-resolution mass results (Supporting information). Additionally, in NOESY spectra of 6b, the clear
NOEs signals have been observed between aromatic (Hh), methylene bridge (Hg) and ethoxyl (Hg and Hf) and the guest alkyl chain
protons (Ha to Hd, Fig. 3). This further proved host pillar[5]arenes were encapsulated on the guest molecular thread and the location of
1
the host pillar[5]arene was on the alkyl chain of the guest thread. H NMR results revealed more complicated situations of 6b and 6c.
Couples of new proton resonance signals were observed in these spectra due to the incorporation of longer alkyl chains in the
precursors. As judged from the resonance spectra, the shielding effect of pillar[5]arene was influenced by the alkyl chain length. For
instance, in the case of compound 6b, the highest field proton signal of the alkyl chain appeared around -2.02 ppm. However, in
compound 6c, the corresponding appeared around -1.51 ppm, whereas, the corresponding protons in 5b and 5c appeared both around
1.90 ppm. In addition, aromatic protons of pillar[5]arene in 6b and 6c exhibited slightly different δ values (corresponding δ for 6b were
6.94 and 6.90; δ for 6c were 6.92 and 6.88). This suggested slightly different shielding effect of pillar[5]arene in synthesized
[3]rotaxanes. In compound 6a, proton Hd was found to be merged with Hb in resonance spectrum, in the case of 6b and 6c, these two
protons clearly split into two separated peaks (δ for Hd in 6b was 7.63; δ for Hd in 6c was 7.40). This suggested varied C-H···O
1
hydrogen bond situations between Hd and oxygen atom in pillar[5]arene in 6b and 6c in contrast to 6a. The detailed H NMR spectra
for 6b, 6c and their corresponding 5b and 5c were shown in Figs. S2 and S3 (Supporting information). We attempted to obtain the
single crystal of the three compounds to conduct an in-depth analysis of the configurations, yet we failed to obtain the desired single
crystals.

Please donot adjust the margins
Fig. 3. NOESY spectrum of 6b in CDCl3.
In order to examine the photo-switch properties of [3]rotaxanes 6a, 6b and 6c, we conducted the light irradiation experiments. As
shown in Figure 4, compound 6a exhibited partially photo-switch property as characterized by UV-vis. Upon irradiation with UV light
of 365 nm with the light intensity of 2.88 mW/cm2 for 30 min, the absorbance of 6a centered at 350 nm, which corresponds to the -*
electronic transition of cis-azobenzene isomer, reduced dramatically. This suggested the occurrence of trans to cis under UV irradiation
of 6a. Subsequently, irradiating 6a with blue light LED resulted in the 350 nm absorption recovery, meaning occurrence of reverse
isomerization [25]. But the absorbance did not completely restore, meaning comprised cis to trans photo-isomerization efficiency.
Compounds 6b and 6c displayed similar incomplete reversible transformation upon light irradiation according to the UV-vis spectra
(Figs. S3 and S4).
1.50
No light
365 nm 30 min
1.25
365 nm 30 min, blue light LED 30 min
1.00
0.75
0.50
0.25
0.00
300
400
500
600
Wavelength (nm)
Fig. 4. UV-visible absorption spectra of 6a (black line); irradiated with 365 nm light (red line) and subsequently irradiated with blue light LED
(blue line). Irradiation time was 30 min.
Irradiating 6a with blue light LED for 30 min resulted in partially regain of the chemical shifts for these protons, indicating the
location regain of the pillar[5]arenes (Fig. 5). The coexistence of both representative chemical shifts for both isomers proved the
incomplete cis/trans isomerization. This is consistent with the result of UV-vis measurement. The detailed cis/trans ratio was
determined by comparing the integration of Hc and Hc’ to be approximately 77/23. Extended time irradiating experiment (150 min of
blue light LED) afforded the identical 1H NMR results as that of 30 min irradiation, indicating the cis to trans photo-isomerization for
compound 6a has reached chemical equilibrium regardless of irradiation time and dose. The further irradiation experiment using the
end capped guest model molecule 5a displayed nearly completely photo-isomerization in both trans-cis and cis-trans directions as
judged by 1H NMR spectra (Fig. S5 in Supporting information). This suggested the installation of pillar[5]arene caused the comprised
trans to cis photo-isomerization efficiency.

Please donot adjust the margins
1
Fig. 5. Partial H NMR spectra of compound 6a upon irradiated with 365 for 30 min and blue light LED, a: no irradiation, b: 365 nm for 30
min, c: blue light LED for 30 min, d: blue light LED for 150 min. (CD2Cl2, 400 MHz).
Irradiation with 365 nm light also resulted in nearly complete trans/cis conversion of compounds 6b and 6c, as proved by
disappearance of corresponding Ha from its original position in 1H resonance spectra. In the backwards photo-isomerization of 6b using
blue light LED, Hd splitted into two peaks in resonance spectra with δ of 7.63 and 7.58, respectively. The peak of δ 7.63 matched the
corresponding peak in 6b without light irradiation. This may suggest a different C-H···O hydrogen bonding situation between Hd and
oxygen atom on pillar[5]arenes in cis and trans isomers of 6b. Also, the cis/trans ratio was determined to be 71/29 by comparing the
intergation of the split Hd. Corresponding Hd in compound 6c (δ of 7.40 without light irradiation) shifted upfield to 7.32 ppm after UV
light irradiation, and partially regained the original chemical shift in 1H NMR resonance spectra. The corresponding cis/trans ratio was
calculated to be 71/29 as well. The similar cis/trans ratio of synthesized [3]rotaxanes indicated the longer alkyl chain of the guest
molecular thread did not result in improved photo-isomerization efficiency. The results of irradiation experiment using corresponding
5b and 5c resembles the case of 5a, further proved the threading of pillar[5]arene could be the main reason for incompletely trans to
cis photo-isomerization (Figs. S7 and S8 in Supporting information).
We subsequently investigated the light-triggered shuttle behaviors of 6b and 6c using 1H NMR. The shielding effects of
pillar[5]arene were observed to increase with alkyl chain length (the most prounced upfield δ variations of the alkyl chain protons of
6b and 6c amounted to -0.07 and -0.09, respectively). The enhanced Δδ of alkyl chain of 6b and 6c in contrast to 6a in the process of
photo irradiation suggested longer alkyl chain could faciliate the shuttle behavior of pillar[5]arene to a certain extent. However, the
photo-triggered shuttle behavior of [3]rotaxanes was believed to be confined to a small extent independent of the alkyl chain lengths.
One plausibile reason could be the strong “hydrogen bond locking effect” between the triazole and pillar[5]arenes that resticted the
1
shifting of pillar[5]arenes even in the streically crowed trans form [3]rotaxanes. The partial H NMR spectra of compounds 6b and 6c
after light irradiation were shown in Figs. S9 and S10 (Supporting information).
Repeated photo-isomerization of compound 6a can be realized by alternatively irradiation with 365 nm light and blue light LED,
respectively. As shown in Fig. 6, the protons in aromatic region of 6a displayed reproducible chemical shift changes within five
irradiation cycles. Compounds 6b and 6c also showed repeated photo-isomerization properties as characterized by alternatively
irradiation experiments. The partial 1H NMR spectra of 6b and 6c were shown in Figs. S11 and S12 (Supporting information).
To best of our knowledge, we have for the first time constructed series of azobenzene bridged pillar[5]arene-based [3]rotaxanes with
1
different alkyl chain guest molecule (6a, 6b and 6c). H NMR and NOESY and high-resolution mass spectra proved the inclusion of
the pillar[5]arenes onto the guest molecule, the main driving force for formation of the [3]rotaxanes include C-H···O hydrogen bond
and C-H···π interactions. As characterized by UV-vis and 1H NMR spectroscopy, the synthesized rotaxanes could undergo completely
trans to cis isomerization under irradiation with 365 nm light for 30 min. Illumination of rotaxanes by blue light did not yield
completely backward cis to trans isomerization even in extended irradiation time and intensity. The installation of pillar[5]arene was
believed to be the reason for incomplete photo-isomerization. The slightly photo-triggered shuttle behavior of [3]rotaxanes have been
1
observed in the H NMR spectra. We are currently focusing on the developing photo responsive [3]rotaxanes with improved shuttle
behavior.

Please donot adjust the margins
Fig. 6. Partial 1H NMR spectra of compound 6a upon irradiated with 365 nm and blue light LED for 30 min for 6 cycles (CD2Cl2, 400 MHz).
Protons in the red squares referred to the aromatic protons.
Acknowledgments
This work was supported by the National Natural Science Foundation of China (Nos. 21372192, 21871227), the Natural Science
Foundation of Jiangsu Province (No. BK20190905) and the Priority Academic Program Development of Jiangsu Higher Education
Institutions. Yang Wang acknowledges the Thousand Talents Plan for Young Professionals of China.
References
[1] (a). T. Kakuta, T. Yamagishi, T. Ogoshi, Acc. Chem. Res. 51 (2018) 1656-1666.
(b). M. Zhang, X. Yan, F. Huang, Z. Niu, H.W. Gibson, Acc. Chem. Res. 47 (2014) 1995-2005.
[2] K. Miyamae, M. Nakahata, Y. Takashima, et al., Angew. Chem. Int. Ed. 54 (2015) 8984-8987.
[3] X.L. Ni, S. Chen, Y. Yang, et al., J. Am. Chem. Soc. 138 (2016) 6177-6183.
[4] M. Zhang, X. Yan, F. Huang, et al., Acc. Chem. Res. 47 (2014) 1995-2005.
[5] (a). Y. Liu, K. Shi, D. Ma, Chem. Asian J. 14 (2019) 307-312.
(b). Y. Sun, J. Ma, F. Zhang, et al., Nat. Commun. 8 (2017) 260.
(c). D. Xia, G. Yu, J. Li, et al., Chem. Commun. 50 (2014) 3606-3608.
(d). G. Yu, C. Han, Z. Zhang, et al., J. Am. Chem. Soc. 134 (2012) 8711-8717.
[6] (a). P. Li, Y. Chen, Y. Liu, Chin. Chem. Lett. 30 (2019) 1190-1197.
(b) Q. Zhang, D.H. Qu, H. Tian, Adv. Opt. Mater. 7 (2019) 1900033.
(c) Y. Zhou, K. Jie, F. Huang, Chem. Commun. 54 (2018) 12856-12859.
(d). Y.M. Zhang, N.Y. Zhang, K. Xiao, et al., Angew. Chem. Int. Ed. 57 (2018) 8649-8653.
[7] D.H. Qu, Q.C. Wang, Q.W. Zhang, et al., Chem. Rev. 115 (2015) 7543-7588.
[8] M. Xue, Y. Yang, X. Chi, et al., Chem. Rev. 115 (2015) 7398-7501.
[9] G. Yu, X. Zhou, Z. Zhang, et al., J. Am. Chem. Soc. 134 (2012) 19489-19497.
[10] R. Zhang, X. Yan, H. Gao, et al., Chem. Commun. Advanced Article, DOI:10.1039/c9cc09155f.
[11] (a). M. Wolf, A. Ogawa, M. Bechtold, et al., Chem. Sci. 10 (2019) 3846-3853.
(b). X. Yao, T. Li, J. Wang, et al., Adv. Opt. Mater. 4 (2016) 1322-1349.
[12] H.Y. Zhou, Y. Han, Q. Shi, et al., Eur. J. Org. Chem. 21 (2019) 3406-3411.
[13] G. Baggi, L. Casimiro, M. Baroncini, et al., Chem. Sci. 10 (2019) 5104-5113.
[14] A. Saura-Sanmartin, A. Martinez-Cuezva, A. Pastor, et al., Org. Biomol. Chem. 16 (2018) 6980-6987.
[15] (a). G. Yu, C. Han, Z. Zhang, et al., J. Am. Chem. Soc. 134 (2012) 8711-8717.
(b). Y. Chen, S. Sun, Y. Yao, et al., Chin. Chem. Lett. 30 (2019) 37-43.
(c). R. Zhang, C. Wang, Y. Yao, et al., Front. Chem. 7 (2019) 508.
(d). S. Sun, T. Liu, Y. Yao, et al., J. Colloid. Interf. Sci. 533 (2019) 42-46.
[16] (a). K. Iwaso, Y. Takashima, A. Harada, Nat. Chem. 8 (2016) 626-633.
(b). Y. Han, C.Y. Nie, S. Jiang, et al., Chin. Chem. Lett. 31 (2020) 725-728.
(c). J. Ye, R. Zhang, W. Yang, et al., Chin. Chem. Lett. 31 (2020) 1550-1553.
(d). R. Zhang, C. Wang, J. Sun, et al., Chin. J. Org. Chem. 39 (2019) 3483-3489.
(e). L.L. Zhao, Y. Han, C.G. Yan, Chin. Chem. Lett. 31 (2020) 81-83.
[17] Z. Yang, X. Liu, S. Zhao, J. He, Prog. Chem. 26 (2014) 1899-1913.
[18] (a). X. Chi, X. Ji, D. Xia, et al., J. Am. Chem. Soc. 137 (2015) 1440-1443.
(b). X.Y. Hu, K. Jia, Y. Cao, et al., Chem. Eur. J. 21 (2015) 1208-1220.
(c). H. Zhang, N.L. Strutt, R.S. Stoll, et al., Chem. Commun. 47 (2011) 11420-11422.
(d). T. Ogoshi, D. Yamafuji, T. Aoki, et al., J. Org. Chem. 76, (2011) 9497-9503.
[19] (a) T. Xiao, L. Zhou, L. Xu, et al., Chin. Chem. Lett. 30 (2019) 271-276.
(b) T. Xiao, W. Zhong, L. Zhou, et al. Chin. Chem. Lett. 30 (2019) 31-36.

Please donot adjust the margins
(c) C. Ke, N.L. Strutt, H. Li, et al., J. Am. Chem. Soc. 135 (2013) 17019-17030.
[20] (a) T. Xiao, L. Qi, W. Zhong, et al., Mater. Chem. Front. 3 (2019) 1973-1993.
(b) D. Cao, H. Meier, Isr. J. Chem. 58 (2018) 1188-1193.
(c) S. Pan, M. Ni, B. Mu, et al., Adv. Funct. Mater. 25 (2015) 3571-3580.
[21] H. Zhang, Z. Liu, Y. Zhao, Chem. Soc. Rev. 47 (2018) 5491-5528.
[22] S. Wang, X. Shao, W. Cai, J. Phys. Chem. C 121 (2017) 25547-25553.
[23] X. Shu, S. Chen, J. Li, et al., Chem. Commun. 48 (2012) 2967-2969.
[24] H. Murakami, A. Kawabuchi, R. Matsumoto, et al., J. Am. Chem. Soc. 127 (2005) 15891-15899.
[25] T. Ogoshi, K. Yoshikoshi, T. Aoki, et al., Chem. Commun. 49 (2013) 8785-8787.