Embed Size (px)
Citation preview
at SciVerse ScienceDirect
Polymer 54 (2013) 4351e4382
Contents lists available
Polymer
journal homepage: www.elsevier .com/locate/polymer
Feature article
New design strategies for second-order nonlinear optical polymersand dendrimers
Wenbo Wu, Jingui Qin, Zhen Li*
Department of Chemistry, Hubei Key Lab on Organic and Polymeric Opto-Electronic Materials, Wuhan University, Wuhan 430072, China
a r t i c l e i n f o
Article history:Received 20 February 2013Received in revised form3 May 2013Accepted 15 May 2013Available online 25 May 2013
Keywords:Nonlinear opticalSuitable isolation groupDendritic structure
* Corresponding author. Fax: þ86 27 68756757.E-mail addresses: [email protected], lichemlab@
0032-3861� 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.polymer.2013.05.039
Open access under CC B
a b s t r a c t
The development of organic/polymeric materials with second-order nonlinear optical (NLO) perfor-mance became more and more important, due to the increasing potential for the applications in photonicdevices and many advantages over conventional inorganic crystalline materials in the last few decades,especially after 1980s. So far, the NLO polymers have developed from the initial guest-host systems, toside-chain polymers, cross-linked systems, dendrimers and dendronized polymers, and then to hyper-branched polymers and even self-assembly systems. As a result, the NLO coefficients have been improvedfrom lower than 10 pm/V to above 300 pm/V. In this article, we would like to review the development ofNLO polymers, especially some recently reported design strategies, such as “site-isolation principle”, theconcept of “suitable isolation group”, the special effect of “isolation chromophore”, AreArF self-assemblyeffect, etc.� 2013 Elsevier Ltd. Open access under CC BY-NC-ND license.
1. Introduction
In 1960, the first ruby laser was born in USA, which is consideredas one of the most significant scientific achievements in the 20thcentury. Soon afterwards, in 1961, second-harmonic generation(SHG) was firstly observed by Franken when one beam of laserpassed through the quartz crystal (a-SiO2), giving the birth to a newresearch field of nonlinear optics (NLO) [1]. Since then, more andmore NLO phenomena, such as photomixing, higher harmonic,Pockels effect, photorefractive, two-photon absorption effect, etc.,have been observed, thanks to the efforts of scientists.
The 21st century is an age of global information. With the rapiddevelopment of information technology, traditionalmicroelectronicmaterials,whichusually use electrons as the carriers, is nowdifficultto meet the requirements for future communication technology.In comparison with that by electrons, there are many advantagesto transmit information by photonics, such as good parallelism,high frequency, wide bandwidth, high speed, anti-jamming andso on. [2]. Thus, one of the crucial limitations in photonics tech-nology is how to prepare electro-optical (EO) materials, one type of
163.com (Z. Li).
Y-NC-ND license.
second-order NLO material, with excellent performance, and thisfield has already become very important in the frontier of materialscience and optical engineering. In earlier research, the majority ofEO materials were inorganic second-order NLO crystals, includingLiNbO3, GaAs [3], some of which have been already prepared on anindustrial scale. However, their lower EO coefficient, higher dielec-tric constant and half-wave voltage, make themdifficult tomeet therequirements of electro-optic modulator with large area [4].Therefore, more and more scientists turned their attention to theorganic materials, since organic materials usually offer potentialadvantages over inorganic crystals, such as low dielectric constants,large and ultrafast responses, facile fabrication and good process-ability, andwide range of operating frequencies. In earlier 1980s, theproposal of “poled polymer” by Meredith brought the organic/polymeric NLO material to a higher level [5], leading a vigorousdevelopment in NLO field. Nowadays, nearly all the organic/poly-meric NLO materials are “poled polymers”.
For chemists, the general approach to improve the performanceof materials is to modify their structure at the molecular level. Andin these years, lots of NLO polymers with different chemicalstructure have been designed to investigate the structure-propertyrelationships of organic NLO materials, and much better perfor-mancewere achieved. In this article, wewould like to present somenew design strategies for second-order NLO materials, especiallypoled polymers.
Fig. 1. Graphical illustration of poling procedure for NLO polymers.
W. Wu et al. / Polymer 54 (2013) 4351e43824352
2. The concept of “poled polymer” and site-isolation principle
To be macroscopic second-order NLO active, the arrangement ofdipoles of NLO chromophores should be noncentrosymmetric.However, as shown in Chart 1 [6], nearly all of the NLO chromo-phores have the structure of a donorepeacceptor (DepeA), whichshould bring out the terrible charge separation. And these elec-trostatic interactions between donors and acceptors could makethe centrosymmetric alignment of the NLO chromophore moieties.Thus, to achieve their noncentrosymmetric alignment, the polingprocedure should be conducted (Fig. 1) [7]: at first, a strong directcurrent (DC) electric field is applied to induce the reorientation ofdipolar NLO chromophores in solid solutions when the materialsare heated near or above the glass transition temperature (Tg); andthen, the materials are cooled to room temperature while thepoling field is still applied, thus, the noncentrosymmetric align-ment of the dipolar NLO chromophores could be “frozen in” to formthe so called “poled polymers” [8]. However, generally, the NLOchromophores have the structures of a donorepeacceptor (DepeA), which usually possess a rod-like structure (Chart 1), resultingin strong intermolecular dipoleedipole interactions in the poly-meric system, and further making the poling-induced non-centrosymmetric alignment of chromophores a daunting task [9].Furthermore, these intermolecular dipoleedipole interactionswould become stronger and stronger, accompanying with theincreased concentration of NLO chromophore moieties in the
Fig. 2. Variation of EO activity with chromophore number density for composite mate
materials, which would finally lead to a decreased NLO coefficient(Fig. 2) [10]. Thus, how to decease these interactions, and efficientlytranslate the large b values (first polarizability, which coulddemonstrate the microscopic second-order NLO activities of NLOchromophores) of the organic chromophores into highmacroscopicNLO activities of polymers, has become one of the major problemsencountered in optimizing organic NLO materials.
Fortunately, things were changed around 2000. The introduc-tion of some isolation groups (IG) to the chromophore moieties tofurther control the shape of the chromophore should be an efficientapproach to minimize these interactions and enhance the polingefficiency. In particular, a spherical shape, proposed by Dalton et al.[11], has been considered as the ideal conformation. This was afamous principle, site-isolation principle, in the NLO research field.It should be the most convenient way to decrease the interactionthrough the enlargement of their distance; and the spherical shapecould further help the chromophores being orderly aligned, just asshown in Fig. 3. The dendritic chromophore C2, which wasencapsulated by three highly fluorinated dendrons, demonstratedlarge NLO effect, and its r33 value was measured to be 30 pm/V,three times that of its prototype C1 (with an r33 value of 10 pm/V)[12]. Here, the r33 value is the EO coefficient, an importantparameter to express the second-order NLO effect. In this article,there is another parameter, SHG coefficient (d33 value), also used toreflect the second-order NLO effect. There is an almost fixed rela-tionship between these two parameters: the tested r33 value wasabout 0.6 of the corresponding d33 ones. Besides the enhanced EOcoefficient, the stability and optical loss, which were other impor-tant parameters for the real applications, of dendritic chromophoreC2, were both higher than those of C1. Applying the site-isolationprinciple, in the past decade, various series of NLO dendrimersand dendronized polymers, which have highly three-dimensionalbranched structure with some bulky IGs linked to chromophoremoieties as the isolation groups, have been prepared and exhibitedlarge NLO coefficients [13], as well as some other required auxiliaryproperties, including optical transparency, chemical and photo-chemical stability and so on [14]. Chart 2 showed some dendron-ized polymers as typical examples [15]. All of these dendronizedpolymers demonstrated large NLO coefficients, and the r33 value of111 pm/V was a breakthrough at that time (earlier 2000 s).
The normal IGs could be also changed to functional groups,which could be cross-linked after or during the poling procedure,with the aim to improve the stability of the achieved NLO effect. Forexample, a series of cross-linkable NLO chromophores C3eC5(Chart 3) functionalized with multi-diene-containing dendrons,were developed by Jen et al. [16a]. The chromophores C3eC5/tris-maleimide (TMI) systems could be cross-linked after the polingprocess, and the cross-linked C3eC5 showed large EO coefficients
rials consisting of the FTC chromophores in poly(methyl methacrylate) (PMMA).
Weaker dipole-dipole interactions
Lager NLO coefficients
Rod-like structure
Spherical structure
N OOOO
O
O
O
O
FF
F
FF
FF
F
FF
FF
FF
F
FF
F
FF
S
O O
O
O
FF
FF
F
F
FF
FF
NC
NC CNNC
N
S
NC
NC CNNC
Chromophore C1
Dendritic Chromophore C2
r33 = 10 pm/V
r33 = 30 pm/V
Fig. 3. A two-dimensional graphical illustration of site-isolation principle and a typical example.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4353
r33 values of 63e99 pm/V, much higher than those obtained fromguest-host EO polymers or uncross-linked chromophore withsimilar structure. Furthermore, over 90% of the original NLO ac-tivities could be retained at 100 �C for more than 500 h. Chromo-phore C6/C7 or C6/tris-acrylate (TAC) was another example ofcross-linkable NLO materials, and their NLO coefficients and sta-bilities were both improved through the crosslinking process [16b].Interestingly, this crosslinking method could be also applied tohost/guest systems. Chart 4 showed the chemical structures of thehost polymer P4, guest chromophore C11, and maleimido cross-linkable chromophores C8eC10 [16b]. After poling, P4/C8eC10systems could be cross-linked, and act as host matrix, while chro-mophore C11 used as the guest chromophore to further improvethe concentration of NLO chromophore moieties. As a result, bothultrahigh EO activity (r33 value >260 pm/V at 1310 nm) and goodalignment stability (at 85 �C) were achieved, simultaneously.
Fig. 4. The 3D space-fill model of different types of chromo
3. The concept of “suitable isolation group”
Although the IGs could bring many advantages into NLO poly-mers as mentioned above, it was conceivable that the NLO perfor-mance could not be always enhanced, accompanying with theincreasing size of IGs. The IGs were non-polar, which could notcontribute to the NLO coefficient directly, but decrease the con-centration of NLO chromophore moieties. According to the one-dimensional rigid orientation gas model [17]:
d33 ¼ 12Nbf 2uðf uÞ2
Dcos3 q
E(1)
where N is the number density of the chromophore, b is its firsthyperpolarizability, f is the local field factor, 2u is the double fre-quency of the laser, u is its fundamental frequency, and <cos3 q> is
phores optimized via the PM6 semi-empirical method.
N
NN
NO2
DR
NHO OH
NO
O
APII
NOAc
NO
OS
ISX
NOAc
S
FCN
AcO
CN
CNC5F11
NOAc
S
AcO
CNNC
NC
CN
TCI
N
S
S
NC
CN
OO
SDS
N
S
AcO OAc
Bu
Bu
O
CN
NCCN
CLD
NO O
Si Si
OCN
NC
CN
FTC
Donor
Bridge
Acceptor
rod-like structure
Chart 1. Some common NLO chromophores.
0.1 0.2
10
20
30
40
50
60
70
80
0.40 0.48 0.56
100
120
140
160
180
200
0.500 0.525 0.550
210
215
220
225
230
235
240
245
250
d3
3v
alu
es
(p
m/V
)
normal chromophore
A B
normal dendrimer
The loading density of the effective chromophore moieties
C
global-like dendrimer
Fig. 5. The SHG coefficients of normal chromophores, normal dendrimers and global-like dendrimers as a function of the loading concentration of the chromophore moieties(A, normal chromophore C1 doped in PMMA; B, normal dedrimers G1eG5; C, global-like dendriemrs G1-TPAeG3-TPA).
O OO O
O F
FF
OO
1-x x
N3
N
OONC
NCCN
OO
OO
OF
FF
F
F
F
:
N
S
N
P1 P2
O OO O
x y
OO
O OF
FF
F
FF
O OO
CNNC
CN
N OOO O
OOF
FF
FF
F
P3
P1: r33 = 81 pm/V
P2: r33 = 97 pm/V
P3: r33 = 111 pm/V
Chart 2. Some examples of dendritic chromophore or dendronized polymers based site-isolation principle.
W. Wu et al. / Polymer 54 (2013) 4351e43824354
N
S
O
CN
CN
CNF3C
O
O
O
O
Ar
Ar
O
O
O
O
Ar
Ar
N
S
O
CN
CN
CNF3C
O
O
O
O
O
O
O
O
Chromophore C3,C4,C6,C7
Chromophore C5
C3: Ar =
C4: Ar =
C6: Ar =
C7: Ar =
O
OO
O
NN
N
NO
O
O O
O
OO O
O
O
O
O
TMI TAC
Chart 3. Structures of DielseAlder cross-linkable dendronized chromophores C3eC7 and the tris-maleimide (TMI) and tris-acrylate (TAC) crosslinkers.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4355
the average orientation factor of the poled film, the decreasedconcentration of NLO chromophore moieties would decrease theNLO effect. The introduction of IGs to the NLO chromophore moi-eties could decrease their strong intermolecular electronic in-teractions, directly increasing the poling efficiency (<cos3 q>).However, some other disadvantages were caused: the introductionof the IGswould dilute the active concentration of the chromophoremoieties, leading to the reduced N value, thus decreasing the d33value; the IGs would make the chromophore larger, causing adifficult alignment of NLO chromophore, which would decrease thepoling efficiency. Therefore, the role of IGs should be complicated,but not always beneficial to the NLO effect. Based on this consider-ation, we have prepared different kinds of NLO polymers, in which
N
S
O
CN
CN
CNF3C
O
O
O
O O O O0.9 0.1
Host polymer P4
NO
O
O
N
O
O
C8
N
S
F3C
O
C
O
O
N
O
O
N
O
OSi
Si
Gues
Chart 4. Chemical structures of the host polymer P4, guest chrom
the size of the isolation groups in NLO chromophore moietieswas changed from small to large, in order to find out how to choosefavorite IGs to boost the fixed microscopic b value of an NLO chro-mophore to possible higher macroscopic NLO property of polymerscontaining this chromophore efficiently. Some typical exampleswere shown in Table 1, including different polymer backbone,different type of polymer, different chromophore, different linkageposition et al. [8f,18ael] Based on these experimental results, aswellas some theoretical analysis, in 2006, we have proposed a newconcept of “suitable isolation group (SIG)”: for a given chromophoremoiety and a given linkage position, there should be a suitableisolation group present to boost its microscopic b value to possiblehigher macroscopic NLO property efficiently [18l].
O
CN
CN
CN
O
N
O
O
9
N
S
O
CN
CN
CNF3C
O
O
O
NO
O
O
N
O
O
O
O
N
O
O
C10
O
NC NC
CN
F3CS
t chromophore C11
ophore C11, and maleimido crosslink chromophores C8eC10.
Table 1Some examples of suitable isolation groups.
No. Chemical structure SIG (-R group) d33a (pm/V) Reference
P5eP8
n
CH3
NOO
NN
NO2
CHN
OC
HN
O
RP5: R = H
P6: R = Br
P7: R =
P8: R = NC4H9
NC4H9
82.3 [18a]
P9eP12P9: R = H
P10: R = Br
P11: R =
P12 R = NC4H9
n
CH3
NOO
NN
S
CHN
OC
HN
O
R
O OC2H5
63.0 [18a]
P13eP16
P13: R =
n
CH3
NOO
NN
NO2
CHN
OC
HN
O
O(CH2)3 N
NN
R
CH3(CH2)5
P14: R =
O(CH2)4P15: R =
N(CH2)4
P16: R =
74.8 [18b]
P17eP19
N(CH2)2O
RN N
SO
O
(CH2)2O
C
C
O HN
HN
CH3
)
)n
O
NC4H9
P17: R = H,
P18: R =
P19: R =
26.2 [18c]
P20eP22
NC4H9
P20: R = H,
P21: R =
P22: R =
)
)n
N
RN N
NO2
O(CH2)2O CO
HN
H3C
HN C
OO(CH)2
54.1 [18d]
P23eP26
NO
NN
S
CHN
O
OO
(CH2)2
n
P23: R = H
P24: R =
P26: R = NC4H9
P25: R = O
OCO
HN(
)
R
O55.9 [18e]
P27eP29
y
CH3N OO
NN
S
CNH
OC
HN
O
CH3
N OOCHN
OC
HN
O
OOCH2CH2OH
x
CH3N OO
NN
S
CHN
OC
HN
O
n
OOCH2CH2O
Bz =
Np =
Cz = N
CO
R
P27: R = Bz, x = 0.71, y = 0.02;
P28: R = Np, x = 0.70, y = 0.03;
P29: R = Cz, x = 0.62, y = 0.11.
0.27
45.6 [18f]
W. Wu et al. / Polymer 54 (2013) 4351e43824356
Table 1 (continued )
No. Chemical structure SIG (-R group) d33a (pm/V) Reference
P30eP32
x
CH3
NOO
NN
NO2
CHN
OC
HN
O
CH3
N OOCHN
OC
HN
O
y
CH3
N OO
NN
NO2
CHN
OC
HN
O
n
P30: R = Bz, x = 0.02, y = 0.58,
P31: R = Np, x = 0.11, y = 0.48,
P32: R = Cz, x = 0, y = 0.59.
OCH2CH2OH OCH2CH2O C RO
0.41
N58.2 [18g]
P33eP36
ONC4H9
N
N
0.5
0.5
n
NCC O
O
N NN
R
NO2
P34: R =
P36: R =
P33: R = H
P35: R =
O39.5 [18h]
P37eP40
P37: R = CH3(CH2)5
P38: R =
(CH 83.8 [18i]
P41eP44O
105.6 [18j]
P45eP48O
83.8 [18j]
P49eP54
(CH2)3
ro=R
C
O
O(Bz)
CO
O
(Np)
P49: R = H; P50: R = ; P51: R = .
P52: R = H; P53: R = Bz; P54: R = Np.
R1 = H
R2 =
O
O
40.1 [18k]39.4
(continued on next page)
W. Wu et al. / Polymer 54 (2013) 4351e4382 4357
Table 1 (continued )
No. Chemical structure SIG (-R group) d33a (pm/V) Reference
P55eP63
roro=R N
R1
NN
NO2
N
NN
NO2
O
N
NN
SO O
C C C C
(CH2)3 (CH2)3
Cl
x yn
x = 0.25y = 0.75
NN
N
R
R2 R2 R3 R3
NC4H9
P55: R1 = H,
P56: R1 =
P57: R1 =
P58: R2 = H,
P59: R2 =
P60: R2 =
CO
O
CO
O
P61: R2 = H,
P62: R2 =
P63: R2 =
CO
O
CO
O
R1 =
R2 = H
R3 = O
O
130.5 [18l]62.853.2
a SHG coefficient at 1064 nm of the polymers with suitable isolation group.
W. Wu et al. / Polymer 54 (2013) 4351e43824358
With thedevelopment of the concept of SIG, at present, itwas notonly description of the phenomenon, but could also predict therelative size of the suitable isolation group for a certain NLO system[8f]. For example, in our case of P5eP12, the only difference betweenthe two series of the polymers is the different acceptor group: one isnitro groups, another is sulfonyl moiety. And nitro group is astronger acceptor than the sulfonyl one. Then, the intermoleculardipoleedipole interactions between the chromophores containingnitro groups as acceptors should be stronger than those present inthe caseof sulfonylmoieties. Also, under the samepoling conditions,the force for the orderly alignment of the chromophore with sul-fonylmoieties as acceptor groups should beweaker than that for thenitro ones. Therefore, to minimize the interaction of the nitrochromophores to the similar degree as in the case of sulfonyl moi-eties, the IG should be larger. Actually, the SIG for P5eP8 wascarbazole, while that for P9eP12 was phenyl ring. According to this,
NS
O
NC CN
CN
R
OSi
OO
OO
OC12: R =
Chromophores C12-C14
O
Changable Isolation groups
C13: R =
C14: R =
N SC6H13
C6H13
N S
O
O
Chart 5. Some examples of NLO
we can predict that if there were two or more IGs bonded to thenitro-based chromophoremoieties, the bulk of the suitable isolationgroup should be smaller than carbazole; if the acceptorwas strongerthan nitro, the SIG should be larger than carbazole. We will discussthis point later.
Some similar excellent work has been also reported by othergroups. For example, Jen et al. synthesized a series of FTC chro-mophores (FTC chromophore is a type of chromophore with N,N-disubstituted aniline as the donor; thiophene as the p-bridge;tricyanofuran-based (TCF) structure as the acceptor, Chart 1showed a typical example for FTC) with different size of IGs(Chart 5, Chromophores C12eC14) [19a]. The results showed thatthe r33 values of C13 and C14 in the guestehost systems (the hostpolymer was polycarbonate, with a glass transition temperature of205 �C), were nearly 40% higher than that of C12. Dalton et al.synthesized a series of chromophoreswith bithiophene as p-bridge
S O
NCCN
NC
S O
NCCN
NCO
O
N S S O
NCCN
NCOO
OO
O
O
N S S O
NCCN
NCOO
OO
O
O
CF3
C15 C16
C17 C18
chromophores with SIGs.
0.063 0.92
OOH2C
NO
O
OO
S
N
O
O
O
NCNC
CN
CF3
OOCH3
OH2
N
O
O
0.017
O OCH2
0.08
OOH2C
NO
O
O
OO
ON
S
O
NC
CF3NC
NC
P68 P
P7
0.08 0.92
OOH2C
NO
O
O
OOCH3
ON
S
O
NC
CF3NC
NC
OR
P71
C
Chart 7. Chemical structures of den
P73
NH
NN
NO2
HNO
NN
NO2
O CHN
OC
HN
O
Φ = 0
"Orde
Chart 8. The structure of polyurethanes em
NO O
R
OCN
CN
CO H
NHN C
O
n
P64: R = H
P65: R = Br
P66: R =
P67: R =
Chart 6. Some examples of NLO chromophores with SIGs.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4359
(Chart 5, Chromophore C15eC18) [19b]. The IGs in chromophore C7should be the SIGs in C15eC17, with normal TCF as acceptor, and itsr33 value was nearly three times that of C17 with no isolation group.However, if the acceptor was CF3-TCF (chromophore C18), whichcould make the b value of C18 enhanced as twice of C16, the polingefficiency was decreased to a large degree, indicating the IGwas notsuitable any longer. This confirmed that the SIG should be differentfor different chromophores. In 2009, Hsiue et al. prepared a seriesof NLO polymers with 4-(dicyanomethylene)-2-methyl-6-(p-(dimethylamino)styryl)-4H-pyran (DCM) as chromophore moietiesin the side chain (Chart 6) [19c], in which, the phenyl group act asthe SIG, and P66 demonstrated the highest d33 value of 68.7 pm/V.
From 2001, the “click chemistry” reaction has aroused muchinterest among researchers because of its remarkable features,
0.08 0.92
OC
NO
O
OO
SO
NCNC
CN
CF3
OOCH3
0.08 0.92
OOH2C
NO
O
OO
S
N
RO
O
NCNC
CN
CF3
OOCH3
0.88
OR
69P70
2
R = *
O O
O
F
F
F
FF
F
F
F
FF
Fluorinated dendriticmoiety in ploymerP70, P71 and P72
0.04
O OH3 CH2
dronized polymers P68eP72.
P5
n
CH3
NOO
NN
NO2
CHN
OC
HN
O
n
CH3
.31.
r parameter doubled"
Φ = 0.15.
bedded with “H-type” chromophore.
N
NN
NO2
O(CH2)3
N
NN
NO2
O(H2C)3
nR2
R1
P75-P80
P75: R1
= R2
=
C6H13 C6H13
P76: R1
= R2
=
OC6H13
C6H13O
P77: R1
= R2
=
(CH2)4 NNNNNN
C6H13 C6H13
P78: R1
= R2
=
(CH2)4 NNNNNN
OC6H13
C6H13O
P79: R1
= R2
=
C6H13 C6H13
P80: R1
= R2
=
OC6H13
C6H13O
CH2H2CN NNNNN
CH2H2CN NNNNN
Chart 10. The structures of “H-shape” polymers.
Chart 11. The structures of polyurethanes embedded with “H-type” chromophore as well as SIG.
OH OH
N NN N
ArAr
OH
NNAr
Chromophore C19-C23 Chromophore C24-C28
NO2C19 and C24: Ar =
C20 and C25: Ar = CF3
C22 and C27: Ar = F
C23 and C28: Ar = Br
C21 and C26: Ar = Cl
Cl
Chart 12. Some structures of “H-type” chromophores.
CO H
N
CH3
n
NCH2CH2O
NO2
C6H13 C6H13
N
O2N
O2HC2HCCOH
Nd33= 57 pm/V,
λmax =425 nm(in THF).
P74
Chart 9. The structure of polyurethanes embedded with indole-based “H-type”chromophore.
W. Wu et al. / Polymer 54 (2013) 4351e43824360
O O
N NN N
ArAr
CF3P85: Ar =
P86: Ar = F
Cl
N
CF3F3CN
O
O
O
On
P85-P87
P87: Ar = Cl
Chart 13. The structures of polyimides embedded with these “H-type” chromophores.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4361
such as nearly quantitative yields, mild reaction conditions,broad tolerance toward functional groups, low susceptibility toside reactions and simple product purification. By utilizing the“click chemistry” reaction of the DielseAlder cycloadditionreaction in NLO area, in 2006, Jen et al. synthesized a seriesof side-chain dendronized polymers P68eP72 (Chart 7) [19d].
O
O
O
NC6H12
N
NN
R
C6H12
N
NN
R
C6H12Star-type chromop
C29-C32
Chart 14. The structures of “star-ty
O
O
O
O
O
O
O
O
O
R
R
R
R
FF
C33
Chart 15. The structures of “star-type” chro
These polymers demonstrated high r33 values in the range of33e60 pm/V. Due to the advantages of the “click chemistry” re-action, it was convenient to control the linkage position of thechromophore moieties, the concentration of NLO chromophores,the size of isolation group, and the introduction of crosslinkgroup and so on.
O
NCNC
CN
O
NCNC
CN
O
C29: R =
NN
Rhore
C30: R =
C31: R = NO2
C32: R =
NCCN
pe” chromophores C29eC32.
NS
NCCNCN
CN
O
OO=
O
OO
O
OF
F
OF
F
Cross-linkable group
mophore C33 prepared by Jen’s group.
Table 2The contradistinction of polymers P88eP90 containing different shape ofchromophores.
No. Chromophore shape d33a (pm/V) Fb lmax
c (nm)
P89 “Star-type” 115.5 0.25 500P88 “H-type” 64.3 0.14 509P90 Normal rod-like 12.6 0.02 504
a SHG coefficient tested at 1064 nm.b Order parameter.c The maximum absorption of polymers in thin film.
W. Wu et al. / Polymer 54 (2013) 4351e43824362
4. Some applications of site-isolation principle and theconcept of “suitable isolation group”
As mentioned above, the site-isolation principle claimed thatthe chromophore with a rod-like structure was very terrible forthe NLO effect, and the ideal conformation should be sphericalshape. Thus, it was reasonable to change the rod-like structure ofchromophores by the introduction of SIG. Among the numerousworks on this idea, “H-type” chromophore should be a typicalsuccessful example, which was firstly reported in 2006 [18a]. Asshown in Chart 8, the order parameter (F), which could show theorderly alignment of the chromophore moieties in the polymers, ofthe resultant polyurethane embedded with this “H-type” chro-mophores, was more than twice that of its analogue with normalrod-like chromophore. Meanwhile, the maximum absorptionwavelength (lmax) of P73 was 463 nm,10 nm blue-shifted than thatof P5, exhibiting much better optical transparency, which wasbeneficial to the practical applications. Shortly after that, this“H-type” chromophore was further applied into indole-based
O
OO
NSO
CN
NCNC
O
OO
O
N
S
ONC
CNNC
O
OO
N
S
ONC
CNNC
O O
OO
NS
O
OO
N
O
OO
OC34
Chart 16. The structure of six
N
NN
O2N
O
N
N
NN
O2N
O
N NN NO2
O
C6H13C6H13
n
P88:
"Star-Type"
Chromophore
Chart 17. The structures of NLO polymers P82eP
chromophores (Chart 9) [20]. Once again, large NLO coefficient(d33 ¼ 57 pm/V) and good optical transparency (lmax ¼ 425 nm)were achieved synchronously.
Thus, how about to combine two or more “H-type” chromo-phores in one NLO polymer? From 2009, our group synthesizeda series of new polymers, in which the linkage mode betweentwo “H-type” chromophores was shoulder-to-shoulder [21].
O
NCCN
CN
S O
NC
CNCN
N
S
OCN
NCCN
O
OO
NS O
NC
CNCN
O
OO
O
N
S
OCN
NCCN
OH
NS
O
NCCN
CN
C35
C36
-branched chromophore.
P89:
N
NN
NO2
O
N
N N
NO2
O
N
CH3
"H-Type"
Chromophore
C6H13C6H13
n
N
NN
NO2
O (CH2)3n
P90:
Rod-like Chromophore
84 with different shapes of chromophores.
N N
NN
NO2
O
NNN
N NN
NN
NO2
O
NNNN
NN
NO2
O
NNN
N NN
NN
NO2
O
NNN
NN N
NN
NO2
O
NNN N
NN
NO2
O
NNNN
G3-G5
NNN N
N NN N
N NN N
N NN N
N
RRRR RRRR
NN N NNN
N
NN
(CH )ONN
N OCCOO O
NO
(CH )ONN
NO
NCO OCOO
NO
O
G4: R =
NN N NNN
N
N N
(CH )ONN
N OCCOO O
NO
(CH )ONN
NO
NCO OCOO
NO
N
N
NN N
NN
NN
(CH) O
(CH )O
O NO
N
N
NN N
NN
NN
(CH) O
NN
NOC
COO
O
O N
(CH )O
NN
O N
N
CO
OC O
O
O N
G5: R =
N N
N
NO
OO
O O
OG3: R =
NO O
NN
NO2
O
OO
NNN
N OO
NN
NO2
O
O O
NNNN
G1
NO O
NN
NO2
O
OO
NNN
N OO
NN
NO2
O
O O
NNNN
NN
NO2
O
NNN
N OO
NN
NO2
O
O O
NNN
NO O
NN
NO2
O
OO
NNN N
NN
NO2
O
NNNN
G2
Chart 18. The structures of NLO dendrimers G1 to G5.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4363
Table 3NLO results of dendrimers G1eG5.
No. d33a (pm/V) Nb lmax
c (nm) Tonsetd (oC)
G1 100.0 0.402 479 (�)e
G2 108.1 0.488 482 (�)e
G3 122.7 0.520 480 70G4 177.0 0.537 470 100G5 193.1 0.544 470 107
a Second-harmonic generation (SHG) coefficient at 1064 nm.b The loading density of the effective chromophore moieties.c The maximum absorption of polymers in thin film.d The onset temperatures for decays in d33 value.e Not obtained.
W. Wu et al. / Polymer 54 (2013) 4351e43824364
Everywhere in this type of polymers, all the chromophore moietiescould be considered as “H-type” ones. Thus, this type of polymerswas named as “H-shape” polymers (Chart 10). In these polymers,the chromophore moieties could be considered as part of thepolymer backbone, and the dipole moments could be easily ori-ented [22]. These polymers demonstrated even larger NLO co-efficients, with a d33 values in the range of 64.4e94.7 pm/V, thanthose of polymers embedded with “H-type” chromophores.
On the basic of “H-type” chromophore, the concept of “suitableisolation group”was further applied tomodify the subtle structuresof NLO polymers (Chart 11, P81eP84) [23]. The NLO coefficients ofthese four polymers were 118.6 pm/V, 127.7 pm/V, 108.1 pm/V and83.5 pm/V, respectively. This was a very interesting result. Thesefour polymers demonstrated higher d33 values than those of P73eP80 containing “H-type” chromophore, indicating that the intro-duction of isolation group would still contribute to decreasing thedipoleedipole interactions. However, unlike other polymers onlycontaining rod-like chromophore moieties [8f,18ael], the d33 valueof P82 was only increased about 10 pm/V in comparison with thatof P81, disclosing that further modifying the subtle “H” structure
RO
O
N NN
R
NN
N
R
O
NNN
O
O
O
OO
R
O O
NN N
RNN
N
R O
NNN
O
OO
O
O
N
G2
=
R = (CH2)3
N NN NO2
O coreN
=
Normal low generation dendrimer
Chart 19. The sketch map differences of dendrimers and g
could not improve the poling efficiency of the resultant polymers toa large degree, and the reduced loading density of the chromophoremoieties might lead to the decreased d33 values, as confirmed byP83 and P84. Meanwhile, the SIG in these polymers was ethox-yphenyl (Chart 11), much smaller than the carbazolyl group in P8(Table 1), although the chromophore moieties were nearly thesame. It was consistent with our prediction before: since thereweretwo IGs bonded to the chromophore moieties in P81eP84,accordingly, the bulk of the SIG should be smaller than that onlybearing one IG linked onto the chromophore moieties (Table 1). Asa deduction, if more than two IGs were linked to the chromophoremoieties, the suitable isolation should be even smaller.
Lu et al. have also conducted some excellent work on the“H-type” chromophores [24]. They designed and synthesized aseries of chromophores, in which 9,10-dihydroanthracene wasemployed as a molecular framework, and two DepeA units wereincorporated in a nonconjugatedmode and fixed at nearly the sameorientation (Chart 12) [24a]. The chromophores with two DepeAunits connected in parallel exhibited a remarkable enhancement intheir b values. For example, the b value of chromophore C19 wastested to be 277 � 10�30 esu in THF at 1064 nm, four times that ofC24 (70 � 10�30 esu). In 2008, they further synthesized somefluoro-containing aromatic polyimides embedded with these “H-type” chromophores (Chart 13) [24b]. To our surprise, the d33 valueof P85, only with eCF3 as the acceptor, was as high as 70.2 pm/V.Meanwhile, its lmax was only about 353 nm, demonstrating excel-lent optical transparency, also confirming that this “H-type” chro-mophore should be a simple way to solve the trade-off betweennonlinearity and transparency in the design of NLO materials onceagain.
“Star-type” chromophore was another famous example, whichwas firstly named by Gopalan in 2004 [25]. These chromophores(Chart 14) were synthesized by post-functional azo-couplingreaction, which could make the synthetic procedure much
RO
O
N NN
R
NN
N
R
O
NNN
O
O
O
OO
R
O O
NN N
RNN
N
R O
NNN
O
OO
O
O
R
OO
NNN
RNNN
R
O
N NN
O
OO
O
O
R
O
O NN
NR
N NN
R
O
NN
N
O
O
O
O
O
R
O
O
NN
N
RNN
N
R
O
NNN
O
O
O
O
O
R
O
O
NNN
R
NNN
R
O
NN N
O
O
OO
O
core
G2-TPA
Global-like dendrimer
lobal-like dendrimers (G2 and G2-TPA as an example).
N
NN
NO2
OO
O
OO
O O
Chart 21. The structure of chromophore C37.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4365
convenient and benefit to the real applications. The EO coefficient(r33) of C30 was tested as 25 pm/V at 1550 nm. Furthermore, these“star-type” chromophores were very stable to photochemicaloxidation in ambient light and air, due to the utilization of stableazo group as p-bridge.
Actually, this “star-type” chromophore was even researched byJen earlier. Chromophore C33 (Chart 15) was the first example ofthis type, reported by Jen’s group in 2001 [26a]. The chromophoremoieties with high mb values were in the arms of “star-type”chromophore, which were isolated from each other by the core.Thus, the strong dipoleedipole interactions could be decreased in alarge degree, and the r33 value of the corresponding polymericsystem was up to 60 pm/V. Furthermore, the trifluorovinyl groupscould be cross-linked in the poling procedure at a high tempera-ture, which could improve the stability of the noncentrosymmetricalignment of chromophores (after 1000 h at 85 �C, the r33 value ofits poled film was still up to 54 pm/V, 90% as just poled). After that,more and more “star-type” chromophores have been designed andsynthesized, with success [26].
Very recently, in 2011, Qian and Cui et al. synthesized a six-branched chromophore (C34, Chart 16) via the incorporation ofan FTC type chromophore C36 with phenyl glycerol ester [27a]. Incomparison with its two-branched analogue C35 and originalchromophore C36, nearly all the performance of the six-branchedchromophore was improved, especially, the d33 value at 1064 nm
NN
O2N O
N
NN
N
N
N
N N
NN
NN
R
R
NN NNNNN
NN
(CH2)3O
NN
N OCCOO O
NO2
(CH2)3ONN
NO2
NCO OCOO
NO2
N N
N
NO2
OO
O O
NNN
(CH2)3O
NN
NO2
NCO OCOO
O
O
G1-TPA R=
G2-TPA R=
G3-TPA R
Chart 20. The structures of global-lik
was about 3 times higher (up to 192 pm/V). This six-branchedchromophore could be considered as the development of “star-type” chromophore, with the enlarged amount of arms. Shortlyafterwards, they further introduced different isolation groupsinto this type of chromophore [27b]. However, similar to the case of“H-type” chromophore, the improvement was very limited.
Both “H-type” chromophore and “star-type” chromophorewere useful structure for NLO materials. However, which is better?
N
N
N
NN
NO2
O
NO2
O
N
N
N
N N
NNN
N
R
R
R R
N N NN
N N
(CH2)3O
NN
N OCCOO O
NO2
NO2
N
N
NN N
NN
NN
(CH2 )3 O
(CH2)3O
O2NO
N
N
NN N
NN
NN
(CH2 )3 O N
N
NOC
COO
O
O2N
(CH2)3O
NN
O2N
N
CO
OC O
O
O2N
=
e dendrimers G1-TPA to G3-TPA.
W. Wu et al. / Polymer 54 (2013) 4351e43824366
To answer this question, an NLO polymer P88 (Chart 17), consistingof “star-type” chromophore, and a polymer P89 (Chart 17), bearingan “H-type” chromophorewere synthesized by us [28]. As shown inTable 2, under the same test conditions, the d33 value of P89 was64.3 pm/V, five times that of P90, which only contained the rod-likechromophore moieties. Excitingly, the NLO coefficient of P88, wasfurther improved to 115.5 pm/V. For a better explanation for thisphenomenon, their configurations were optimized via the PM6semi-empirical method, with their 3D space-fill model displayed inFig. 4. The shape of the “star-type” chromophore was much morelike spherical shape, which is ideal to increase the poling efficiency,than the “H-type” one. This should be the main reason for thehigher performance of the “star-type” chromophore.
According to the site-isolation principle, the spherical shapeof chromophore has been considered as the ideal conformation.
N
N
Suitable Core
N
G1-GL
G2-GL: R =
NO O
NN
NO2
O
OO
NO O
N N
NO2
O
OO
NNN
N OO
NN
NO2
O
O O
NNNN
NN
NO2
O
NO O
NN
NO2
O
OO
NNN NNN
NN
NO2
O
G3-GL: R =
G
NO O
NN
NO2
O NNN
N
O
O
NN
O2N
O
N NN
N
O
O NN
NO2O
NNN
O
O
O
O
OO
+
Chart 22. The structures of global-li
Then, if we directly control the conformations of NLO polymers asor similar to the spherical shape, what would happen? High gen-eration dendrimers should be the good candidate. Meanwhile, asmentioned above [23], the more IGs linked to one chromophore,the smaller SIG should be. Thus, if there were three IGs linked toone chromophore, the SIG should be very small. Therefore, wedesigned a series of NLO dendrimers, G1eG5 (Chart 18) [29],through the sharpless “click chemistry” reaction, in which thenewly formed triazole rings during the click chemistry reactionmight act as the SIG to enhance the macroscopic NLO effects. TheNLO effects of these dendrimers (up to 193.1 pm/V for G5) weremuch higher than our other NLO polymers bearing similar azo-benzene chromophores, including those with SIGs. This should beascribed to the dendritic structure of these NLO dendrimers and thegood isolation effect of triazole and benzoic groups. Furthermore,
N
N N
N
NN
N
NN
Too Large Core
NN NNN
N NN
Suitable Core
N
N
NNN
O2N O
NNN
N NN
NN
NO2
ONNN
N
N
N
N N
NO2
O
NNN
N N
NN
NN
N N
NN NN
R
R
R R
R
R
N OO
NN
NO2
O
O O
N
NNN
N OO
NN
NO2
O
O O
NNN
NO O
NN
NO2
O
OO
NNN N
NN
NO2
O
NNNN
NO2
ON
N4-GL: R =
G2-GL to G4-GL
ke dendrimers G1-GL to G4-GL.
O
O
NC
NC
O
OO
O
NC O
O
O
NC
O
CN
O
O
O
CN
O
CN
P91
N
HH
CNNC
O O
ROO
CNH
NR
H
CNO
O
RNO
CNO
H
N R O
O
NC
H
N(CH ) OOCCH CN
H
CNO
O
RN
H
H
H
O
CNO
CNO
O
RN
HO
H
O
H CN
OO
RN
H
O
HO
H
CN
OO R
N
HNC
OO
H
NCO
O
RN
HNC
OO
NC
H
O
O
RN
HO
O
H
RN
HH
CN
O
NC
O
RN
H
CN
O
O
HNC
O O
R N
OH
H
O
RN
O
H H
CN
O
R=-(CH2)11-
P92
Chart 23. The structures of NLO hyperbranched polymers P91 and P92.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4367
as shown in Table 3, accompanying with the growth of NLO den-drimers, the loading density of the effective chromophore moietiesincreased accordingly, from 0.402 for G1 to as high as 0.544 in G5.As mentioned above, from Eq. (1), the d33 value is proportional tothe density of the chromophore moieties. However, due to thesignificantly stronger dipoleedipole interactions, the relationshipbetween the chromophore density and the NLO effect was notlinear any longer, and there was a maximumvalue of NLO effect at acertain concentration of the chromophore moieties. Here, theloading density of the chromophore moieties increased upon thegrowth of NLO dendrimers, correspondingly, the tested NLO effect
O
O
N
NC
NCCN
O
OO
O
OO
N
O
O
O
O
O
CN
CN
CN
P9
Chart 24. The structures of NLO hyper
became larger, demonstrating a deviation from the dipolar frus-tration that typically limits the NLO effect in conventional chro-mophore/polymer compositematerials. More excitingly, from G3 toG4, the d33 value was enhanced 55 pm/V, although the loadingdensity of the chromophore moieties only increased 0.017, whileonly about 14 pm/V from G2 to G3, no matter that the increase ofthe loading density was 0.032. From G4 to G5, the loading densityincreased only 0.007, however, the d33 value was still enhanced16 pm/V. This might indicate that after the loading density of thechromophore moieties increased higher than a critical point,accompanying with the increasing of the loading density, the NLO
O
O
O
O
OHN O
O
O
N
O O
CN
CN
CN
O
O HN
O
O
3
branched oligomeric dopant P93.
N
NN
NO2
OC4H9
N
NN
NO2
N
NN
O2N
C4H9O
P94
Chart 25. The structures of azo-functionalized NLO hyperbranched poly(-aryleneethynylene) P94.
W. Wu et al. / Polymer 54 (2013) 4351e43824368
effect would be enhanced dramatically. Furthermore, accompa-nyingwith the growth of NLO dendrimers, the topological structureof these dendrimers should be closer to ideal, and the other twoimportant parameters (optical transparency and stability for NLOcoefficient) for the NLO materials were all improved, as shown inTable 3. Thus, our results could provide some useful information forthe rational design of NLO materials with better performance.
However, these dendrimers were conical, but not spherical(Chart 19, taking G2 as an example). By utilizing the “star-typechromophore” to further modify the shape of these series of den-drimers to nearly spherical shape, perhaps, even higher macro-scopic NLO effects could be achieved (Chart 19, G2-TPA). Thus, in2012, we attempted to design a new series of global-like den-drimers, G1-TPA to G3-TPA (Chart 20), in which the lower genera-tion NLO dendrimers in our previous work act as the dendron,triphenylamine (TPA) as the core [30a]. To our surprise, the d33value of G1-TPA was already 214.2 pm/V, even higher than that ofG5 (193.1 pm/V). More excitingly, the d33 values increased fromG1-TPA to G3-TPA (246.0 pm/V), accompanying with the increasingloading density of the chromophore moieties, and the higher thegeneration of dendrimer, the higher the d33 value. Fig. 5 showed theSHG coefficients of normal chromophore with SIG (C37, Chart 21),normal dendrimers G1 to G5 and global-like dendrimers as afunction of the loading concentration of the chromophore moieties.It was easily seen that the d33 value of the resultant C37-dopedPMMA films decreased while the loading concentration of thechromophore moieties was only 0.2, much less than the loadingdensity of the chromophore moieties in G1eG5 and G1-TPA to G3-TPA (Fig. 5B, C). However, the NLO coefficients of these dendrimersand global-like dendrimers always increased, accompanying withthe increasing of the loading density. Furthermore, as shown inFig. 5B, C, after the loading density of the chromophore moietiesincreased higher than a critical point (a little higher than 0.5), thed33 values of these dendrimers would be enhanced dramatically. Itwas noted that in comparison with G5, G3-TPA possessed a lowerloading density of chromophore moieties (0.502 vs 0.544), butexhibited a much larger d33 value (246.0 vs 193.1 pm/V). Thisshould be attributed to its three-branched structure (or global-likestructure), and partially realizing our idea of the design of good NLOmaterials by modifying the structure according to the “suitableisolation group”.
For the real application, besides the performance of the mate-rials, the synthetic efficiency and yield were also very important.Due to the low reactivity of the Sonogashira coupling reaction andthe difficult purification, the yields of G1-TPA to G3-TPA were verylow (the total yield of two steps for the core was only 8.99%). Thus,the “click chemistry” reaction, with nearly quantitative yields andsimple product isolation, was used instead of the Sonogashiracoupling one. Meanwhile, to avoid the large isolation effect of new-formed triazoles, the core was changed to phenyl rings but not theprevious TPA units, according to the concept of the “suitableisolation group” (Chart 22) [30b]. As expected, the total yield of thecore through two synthetic steps, was up to 82.9%, thanks to thepowerful “click chemistry” reaction. And the target global-likedendrimers G1-GL to G4-GL (GL is short for global-like) demon-strated similar performances to G1-TPA to G3-TPA, with the d33value of G4-GL up to 233 pm/V. Furthermore, the “click chemistry”reaction was a cycloaddition reaction, and the atom utilizationcould achieve 100%. Thus, through the whole synthetic route, thereshould be only a little byproduct, complying with the requirementsof “green chemistry”.
On the other hand, different from high generation dendrimers,which were generally prepared tedious and costly because of re-petitive protection, deprotection, and purification steps, hyper-branched polymers should be of special interest for their easy
synthetic accessibility, typically by one pot syntheses, which allowsfor their production in large quantities and their application on anindustrial scale. Although the structure of hyperbranched polymersis not perfect enough and still contains linear units, they inherit theproperties of dendrimers such as good solubility, low viscosity, andmultifunctionality at the end groups [31]. It is more important thatthe three-dimensional (3D) spatial separation of the chromophoresendows the polymers with a favorable site-isolation effect, andtheir void-rich topological structure helps to minimize optical lossin the NLO process. Earlier in 1997, Zhang et al. firstly conducted theresearches on NLO hyperbranched polymers. However, the NLOproperties of P91 (Chart 23) was only 2.8 pm/V [32a]. With theintroduced carbazole group, the d33 value of the obtained P92 wasenhanced to 7 pm/V, still very low in comparison with the linearpolymers at that time [32]. Thus, their work did not attract muchinterest. The low d33 values might probably be caused by the low b
value of the chromophore moieties used and the difficulty in theorderly alignment of the NLO chromophoremoieties, whichwere atthe cross positions of the branches, like the head-to-tail polymers.
In 2005, Wang et al. designed some hyperbranched oligomericdopants (P93, Chart 24) with a number average molecular weight(Mn) of 3300 [33]. When its concentration was about 15 wt% inhost polymers, the resultant polymeric system exhibited high NLOeffects, with r33 values up to 65 pm/V, much higher than that oflinear analogue (18 pm/V). Furthermore, the poled film could becross-linked at 200e210 �C, and the r33 value of the obtainedcross-linked poled films could remain about 90% of the originaleffects, even stored at 85 �C for 1200 h under an atmosphere ofnitrogen. In 2004, an ultrahigh d33 value of 177 pm/V was achievedby very simple azo-functionalized hyperbranched poly(-aryleneethynylene) P94 (Chart 25), reported by Tang and Li [34].Furthermore, the stability of its analogue, poly(aryleneethynylene)with TPA moieties instead of the phenyl ring, was very well, andthe onset temperature for decays in d33 value was found to be ashigh as 152 �C.
N
O O
NN
CNO
ON
O
NN
ArH
SO
O
CNO
SO O
CNO
H
Ar
H CN
HAr
CNH
OO
O
O O
O
CN
H
OAr = CZ (P95)
TPA (P96)
Bz (P97)
N
TPA
Bz
NC H
CZ
N
O O
N
N
Ar
SO O
C H
O
CN
H
O
H
n
P98-P100
~ 4 times
N
NN
NO
O
OO
O
N
NO
N
NOO
OO
OO
Br Br
O
NN
NOO
OO
Br Br
NO
P101
N
NN
O
NO
O
Br
O
P102
d33
= 153.9 pm/Vd
33= 98.2 pm/V
N
N
NN
O N O
N
N
NN
O N O
N
O
NO
NN
N
N
NN
O N
O
O
NO
NN
N
N
N
N N
NO
O
P103
d33
= 143.8 pm/V
N
NN
O
NOP104
N
CH
n
d33
= 52.9 pm/V
Chart 26. Some examples of NLO hyperbranched polymers and their linear analogues.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4369
Stimulated by these successful examples, from 2006, weprepared many series of NLO hyperbranched polymers and theirlinear analogues (Chart 26 showed some examples), and all theexperimental data gave one same conclusion (Table 4): the d33values and stabilities of these hyperbranched polymers should be
higher than those of the corresponding linear polymers [35];sometimes, the optical transparency and stability could be alsoimproved transparency and stability could be also improved[35a,b]. These indicated that the 3D architectural structure of thehyperbranched polymers and the spatial chromophore isolation
Table 4Some properties of NLO hyperbranched polymers and their linear analogues.
Hyperbranched polymers no. d33a (pm/V) lmax
b (nm) Tonsetc (�C) Linear polymers no. d33
a (pm/V) lmaxb (nm) Tonset
c (�C)
P95 55.1 407 (�)d P98 13.4 407 (�)d
P96 44.7 434 (�)d P99 20.1 443 (�)d
P97 20.6 432 (�)d P100 (�)d (�)d (�)d
P101 153.9 464 93 P102 98.2 475 70P103 143.8 488 153 P104 52.9 491 119
a Second-harmonic generation (SHG) coefficient at 1064 nm.b The maximum absorption of polymers in THF.c The onset temperatures for decays in d33 value.d Not obtained.
N
N N
NO
O
Ar
Ar
Ar
N
N N
NO
O
Ar
Ar
N
N N
NO
O
Ar
Ar
O
NO
N N
N
ON
NO
N
N
Ar
N
NN
NO O
Ar
Ar
N
Ar
N N
NO
O
Ar
N
N N
NOO
P105-P107
P105: Ar =
P107: Ar =
N
CH3
P106: Ar =
NC4H9
(Bz)
(Car)
(Tri)
P108-P110
P108: Ar =
P110: Ar =
N
CH3
P109: Ar =
NC4H9
N
N N
SO O
Ar Ar
Ar
N
N N
SO O
Ar
Ar
SO
ONN
N
N
NN
SO
O
N
NN
SO
O
Ar
NAr
NN
SO
O
Ar
N
N N
S OO
Ar
(Bz)
(Car)
(Tri)
Chart 27. Some examples of NLO hyperbranched polymers.
Chart 28. The synthesis of AB2-type NLO hyperbranched polymers by “click chemistry”.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4371
in the macromolecular spheres were beneficial to the enhance-ment of their optical nonlinearities. Thus, hyperbranched NLOpolymers might be another choice besides NLO dendrimers anddendronized polymers, to efficiently translate the large molecu-lar first b values of chromophores into high macroscopic NLOcoefficients of materials. In 2012, according to the concept of SIG,by changing the bulky size of comonomers, we synthesized twonew series of hyperbranched main-chain polymers (P105eP110,Chart 27) to investigate the structureeproperty relationship inhyperbranched polymers [36]. The results indicated that even inhyperbranched polymers, the SIG could still work and the d33value of P105 with SIG as comonomer was up to 152.6 pm/V, 2.5times that of P107.
Considering the remarkable features of the “click chemistry”reaction and the good isolation effect of the formed triazole ringsobserved in the above NLO dendrimers, we have designed amodified synthetic approach with the usage of end-capping azomolecules, so that the self-oligomerization of AB2 monomers re-ported previously [37] could be avoided, to produce new azochromophore-containing hyperbranched polymers, P111 and P112,
Chart 29. Illustration of the three possible organizations of the NLO chromophores (arro
from their corresponding AB2 monomers (Chart 28) [38a]. Thesetwo polymers were easily soluble in common organic solvents, andcould form thin films easily. The tested NLO properties demon-strated better macroscopic NLO effects (up to 124.4 pm/V), incomparison with their linear analogue containing the chromo-phores with similar structure. Through the similar syntheticapproach, six new AB2-type polytriazoles containing azo chromo-phore moieties, derived from the same hyperbranched polymerintermediate, were successfully obtained, in which different isola-tion groups in different sizewere introduced to the periphery of thehyperbranched polymers as end-capping moieties [38b]. With thedifferent end-capping groups, these hyperbranched polymersexhibited different solubility and processability; also, their NLOactivities weremodified accordingly, confirming that the concept ofSIG could be applicable in hyperbranched polymers once again.Very recently, to avoid the AB2-type click polymerization, whichwas very difficult to control, the Sonogashira coupling polymeri-zation was also utilized to construct another series of hyper-branched polymers with different end-capped moieties as isolationgroup [38c].
ws) in hyperbranched polymers: (a) main-chain, (b) side-chain, and (c) peripheral.
Chart 31. Schematic overview of different types of macromolecule.
W. Wu et al. / Polymer 54 (2013) 4351e43824372
All the above examples of NLO hyperbranched polymers wereside-chain or main-chain. Besides those, another possible organi-zation of NLO chromophore in hyperbranched polymers was in theperipheral (Chart 29) [39]. In 2009, Odobel et al. synthesized twohyperbranched polymers of this type (Chart 30), which exhibitedrelatively higher d33 values (65 and 42 pm/V, respectively) andgood stability (the d33 coefficients remain stable up to 130 �C) [39].Furthermore, the post-functionalization of the hyperbranchedpolymers provided a new way to attach the NLO chromophores tothe polymers, to rapidly tune the properties of the materials.
Since the 3D topological structure could affect the NLO co-efficients in a large degree, it is very important to design new typesof functional polymers with novel 3D topological structures. Atpresent, our group tried to introduce the dendrimers into anotherdendritic polymeric backbone (for example: hyperbranched poly-mers), in order to develop the types of dendritic structure(Chart 31). Due to their architectural features, these new polymerswere named as “dendronized hyperbranched polymers (DHP)”.Luckily, both the NLO property (with d33 value of 133 pm/V) andstability of the obtained polymers P115 and P116 (Chart 32) wereimproved, in comparison with their analogue of dendronizedpolymers and dendrimers [40]. Further study on this specialdendronized hyperbranched structure is still in progress in ourlaboratory.
5. Aromatic/perfluoroaromatic self-assembly effect in NLOfields
In all of the above cases, the isolation groups used were non-polar groups to decrease the strong interactions between the
Chart 30. The structures of hyperbra
chromophoremoieties, according to site-isolation principle and theconcept of “suitable isolation group”. Actually, the interactionsbetween two isolation groups must be also considered to designthe NLO materials. Different from normal aromatic rings, per-fluoroaromatic rings were electropositive, and this activity couldlead to the reversible self-assembly of these two types of aromaticrings. For example, Patrick and Prosser first recognized that a 1:1mixture of benzene (with a melting point of 5.5 �C) and hexa-fluorobenzene (with a melting point of 4 �C) could form a complex,which melts at 24 �C, much higher than 5.5 or 4 �C [41]. Althoughpure benzene adopts an edge-to-face structure in the solid-state,
nched polymers P113 and P114.
N
Br Br
+ N
B B
B
OO
O O
OO
Pd[PPh3]4, K2CO3
THF/H2O
R
N
N
N
N
N
R
R
R
NN
R
N
NR
P115: R =
P115 and P116
NO O
NN
NO2
O
OO
NNN
NO O
N N
NO2
O
OO
NNN
N OO
NN
NO2
O
O O
NNNN
NN
NO2
O
NNN
P116: R =
"Dendronized Hyperbranched Polymers" (DHP)
Chart 32. The synthesis of “dendronized hyperbranched polymers” P115 and P116.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4373
it has been determined that the structure of the benzene/hexa-fluorobenzene mixture consists of alternating stacks of these twomolecules, due to the electropositive activity of hexafluorobenzene(Chart 33) [8d,42].
Recently, this AreArF self-assembly effect was used in nonlinearoptical (NLO) materials, firstly by Jen et al. in 2007 [43]. By utilizingaromatic/perfluoroaromatic dendron-substituted NLO chromo-phores through the presence of complementary these AreArF in-teractions, they developed a new class of molecular glasses (Chart34) [43a]. Due to this AreArF interactions (Chart 35), the r33 valueof chromophore C40 (108 pm/V) was two times higher than that of
Chart 33. Different interactions bet
chromophore C38 (51 pm/V) and C39 (52 pm/V). More excitingly,this molecule could be also used as host materials in the binarychromophore system, and an ultrahigh r33 value of 327 pm/V wasachieved in the system (C40/C41 ¼ 1:1). Two years later, theyobserved the same phenomenon again in another series of den-dritic NLO chromophores C42eC44 containing perfluoroaromaticas the dendron (Chart 36) [14]. In comparison with the conven-tional chromophore-polymer composite systems, the resultantdendritic chromophores exhibited ultra-large EO coefficients(249e318 pm/V at 1310 nm), improved thermal and temporalstability, and a blue-shifted near-IR absorption tailing. Meanwhile,
ween different aromatic rings.
N
S
OF3C CN
CNNC
O
O
O
OO
O O
O
F
FF
F
F
FF
FF
F
N
S
OF3C CN
CNNC
O
O
O
OO
O O
O
N
S
OF3C CN
CNNC
O
O
O
OO
O O
O
F
FF
F
F
FF
FF
F
F
FF
F
F
F
F
FF
F
N
Si
Si
O
NC
CN
CN
F3CS
C38 C39
C40
C41
Chart 34. Some NLO chromophores containing different types of isolation groups.
Isolation group
Self-assembly effect
NLO chromophore
Guest chromophore
A B
Chart 35. Graphical illustration of the alignment formation of self-assembled chromophore C40 (A) and self-assembled chromophore C40 lattice doped with a secondary chro-mophore C41 (B) after poling.
N O
O
OCN
NCCN
F3C
O O
O
O
O
Ar
Ar
O
O
FF
F
FF
F
FF
F
F
Ar=
C42 C43 C44
C42-C44
Chart 36. Some NLO chromophores containing perfluoroaromatic rings as isolation groups.
W. Wu et al. / Polymer 54 (2013) 4351e43824374
N
NN
NO2
OO
O
Ar2 nP117: Ar
1=
Ar2=
P118: Ar1= P119: Ar
1=
Ar2=
P120: Ar1=
F
F F
F
FF F
F
FF
NAr
2=
NAr
2=Ar1
Chromophore
Self-Assembly
Chart 37. NLO polymers P117eP120 containing different types of isolation groups.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4375
an increase of the EO activity accompanying with the chromophorenumber density was observed even at such high densities for thesechromophore moieties, indicating the advantages of this self-assembly effect in NLO filed in some sense.
Considering the above-mentioned powerful AreArF self-assembly effect, we were wondering if this effect could becoupled with the concept of SIG, to further improve the NLO per-formance of the resultant polymers. From this standpoint, four newsimple NLO polyaryleneethynylenes P117eP120 (Chart 37) con-taining nitro-based azobenzene as the chromophore were designedand prepared successfully in 2012 [44a]. Meanwhile, some smallmodel molecules with similar structure to these polymers werealso prepared, to aid the investigation of the AreArF self-assemblyeffect conveniently. From their NMR spectra and some otherphysical properties, we could know that in this case, the self-
CuIPPh
Pd[PP
THF/E
N
NN
NO2
OO
O
Ar1
+N
Br Br
Br
IOC4H9
I
I
or
P121: Ar1=
Ar=
P122:Ar1= P123:Ar
1=
Ar=
P124:A
F F
F
FF
NAr=
N
OC4H9
Chart 38. NLO hyperbranched polymers P121eP124
Table 5The NLO hyperbranched polymers and their linear analogues.
No. Type ofIG
d33a (pm/
V)Tonset
b
(�C)No. Type of IG d33
a (pm/V)
Tonsetb
(�C)
P117 Phenyl 21.4 90 P118 Pentafluorophenyl 128.5 104P119 Phenyl 50.4 64 P120 Pentafluorophenyl 166.7 74
a Second-harmonic generation (SHG) coefficient at 1064 nm.b The onset temperatures for decays in d33 value.
assembly was present between the isolation groups and the poly-mermain-chain, and would not affect the chromophoremoieties. Itwas exciting that the NLO effect and stability of P118 and P120 wereenhanced in a large degree, in comparison with their analogueswithout this self-assembly effect. The NLO coefficient of P120reached up to 166.7 pm/V, which was, to the best of our knowledge,the new record reported so far for linear polymers containingsimple azobenzene chromophoremoieties. Furthermore, to destroythe alignment of the chromophore moieties in P118 and P120,much more energy should be needed than in P117 and P119, sincethe AreArF self-assembly effect was still present after poling. Thus,the AreArF self-assembly effect in NLO systems could also increasethe stability of NLO polymers (Table 5), which could contribute totheir practical application in the photonics field. We also used thisAreArF self-assembly effect in the design of hyperbranched poly-mers (Chart 38) [44b]. Interestingly, for different comonomers, theself-assembly behavior of P122 and P124 was different, leading totheir different NLO activities. In comparison with P121 containingnormal phenyl rings as isolation groups or its correspondingchromophore, there were many evidences such as much higherglass transition temperature, different NMR spectra, to confirm thepresence of the self-assembly in P122. And these interactionsresulted in much higher NLO coefficient (up to 78.9 pm/V), twicehigher than that of P121 (38.1 pm/V). However, the self-assemblyeffect did not exist in P124, possibly caused by its too small
3h3]4t3N
Ar
N
N
Ar
NN
Ar
NN
N
NN
NO2
O2N
= polymer branch
NO2
O
O
Ar1 O
Ar
OO
Ar1
O
P121-P124
O
O
O
Ar1
r1=
F
F F
F
F
Ar=
OC4H9
containing different types of isolation groups.
N
NN
NO2
O
NO
O
N
NN
NO2OOO
N
NN
O2N O
O O
N
Self-Assembly
N
NN
NO2
O
O O
OC4H9
N
NN
NO2O
O
O
N
NN
O2N
OO
O
C4H9O
=
P122
P124
=
FF
F
F F
Chart 39. Different self-assembly behaviors of AreArF in P122 and P124.
W. Wu et al. / Polymer 54 (2013) 4351e43824376
comonomer unit, which might not supply enough normal phenylrings for the self-assembly effect (Chart 39). On the contrary, theremight be some interactions between the more electropositivepentafluoroaromatic rings and other aromatic ones, which wouldperhaps destroy the comparative perfect 3D structure of hyper-branched polymers to some extent, and lead to a lower d33 value.
Considering the ultrahigh d33 value of high generation den-drimers and global-like dendrimers as mentioned above [29,30], itwas reasonable to wonder if this AreArF self-assembly effect couldstill work in high generation dendrimers, perhaps, much betterNLO effect could be achieved. However, it was known that highgeneration dendrimers were always prepared tediously and
N
NN
N
N
NN
R
O
O
Ar O
Ar
O
P125-P128
Chart 40. NLO hyperbranched polymers P125eP128
expensively because of repetitive protection, deprotection, andpurification steps. Thus, it would be unwise to modify theirchemical structures of dendrimers directly without any preliminaryexperience. Thus, at first, we used the hyperbranched polymers,which have the similar topological structure as dendrimers butvery easy to be prepared in comparison with dendrimers, as themodels. We synthesized a series of azo chromophore-containingAB2-type hyperbranched polytriazoles (P125eP128, Chart 40)bearing perfluoroaromatic rings in different part of these hyper-branched polymers [45a]. The results showed that the penta-fluorophenyl in the periphery produced higher d33 values than thenormal phenyl in the periphery, whereas the perfluoroaromatic
NNN
N
N
O
O
ArO
ArO
O
OAr
O
ArO
NN
NO2
=(CH2)3
=rA=R:521P
=rA=R:621P
=rA=R:721P
=rA=R:821P
(CH )
(CH ) OO
F
F
F
F
(CH )
F F
F
FF
(CH ) OO
F
F
F
F
F F
F
FF
containing different types of isolation groups.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4377
rings in the interior architecture produced slightly lower d33 values.Inspired by this result, in 2013, we designed and synthesized a newseries of NLO dendrimers, G1-PFPh to G5-PFPh (Chart 41, PFPhmeans their end-capped groups were pentafluorophenyl moietiesto differentiate from G1 to G5) [45b], and a new series of NLOglobal-like dendrimers, G1-PFPh-GL to G4-PFPh-GL (their structurewas very similar as G1-GL to G4-GL, only using pentafluorophenylgroups instead of normal phenyl groups in the periphery) [30b].
N N
NN
NO2
O
NNN
N NN
NN
NO2
O
NNNN
NN
NO2
O
NNN
NNN N
N NN N
RRRR
NO O
NN
NO2
O
OO
NNN NN
G1-PFPh
NO O
NN
NO2
O
OO
NNN
N OO
NN
NO2
O
O O
NNNN
NN
NO2
O
NNNN
G2-PFPh
FF
FF
F
FFF
FF
FF
FF
F
FF
F
FF
FF
F
FF
FF
F
F FF
F
FF
G3-PFPh
G3-PFPh: R =
NO O
NN
NO2
O
OO
FF
FF
F F
FF
FF
NO O
N N
NO2
O
OO
NNN
N OO
NN
NO2
O
O O
NNNN
NN
NO2
O
FFF
F F
FF
FFF
FFF
F F
F F FF
F
G4-PFPh: R =
NO O
NN
NO2
O
OO
NNN
FFFF
F
F F FF
F
G
Chart 41. The structures of NLO de
The result was encouraging, similar to G1eG5, while the loadingconcentration of the chromophore moieties changing from 0.326 inG1-PFPh to 0.411 in G2-PFPh, then to 0.444 in G3-PFPh, to 0.459 inG4-PFPh, to 0.466 in G5-PFPh, the measured NLO coefficient valuesincreased from 84 (G1-PFPh) to 89 pm/V (G2-PFPh), then to108 pm/V (G3-PFPh), then to 187 (G4-PFPh) and 206 pm/V(G5-PFPh), which was higher than that of G5 (193.1 pm/V), andthe d33 value of G4-PFPh-GL was further enhanced to 252 pm/V
N NN
NN
NO2
O
NNN
NN N
NN
NO2
O
NNN N
NN
NO2
O
NNNN
N NN N
N NN N
N
RRR R
N OO
NN
NO2
O
O O
NN
N OO
NN
NO2
O
O O
NNN
NO O
NN
NO2
O
OO
NNN N
NN
NO2
O
NNN
FF
FF
F FF
F
FF
FF
F
FF
F FF
F
FF F
F
FF
F
to G5-PFPh
N OO
NN
NO2
O
O O
NNNN
NN
NO2
ONNN
N OO
NN
NO2
O
O O
NNN
NO O
NN
NO2
O
OO
NNN N
NN
NO2
ONNN
N
FFF
FF F F F
FF FFF F
F F FFFF
FFF
FF
F FFFF
NO2
ON
N5-PFPh: R =
ndrimers G1-PFPh to G5-PFPh.
W. Wu et al. / Polymer 54 (2013) 4351e43824378
(for G4-GL, the d33 value was 233 pm/V) [30b]. Meanwhile, theoptical transparency of these dendrimers was also improved in alarge degree, thanks to the powerful AreArF self-assembly effect. Assimilar to linear polymers, the self-assembly effect in dendrimerscould also increase the stability of NLO coefficient. For example, thetemperature for the decay in d33 value of G4-PFPh-GL was up to106 �C, higher than that of G4-GL (94 �C).
6. Isolation chromophore: the expansion of isolation group
In all of the previous cases, the isolation groups were non-polargroups to decrease the strong interactions between the chromo-phore moieties, which did not directly contribute to the macro-scopic NLO effect but decrease the effective concentration of theNLO chromophore moieties, including pentafluorophenyl groups.Then, how about use one chromophore with lower mb value asisolation group for another chromophore group with high mb one,to achieve high poling efficiency, as the result of the decreasedstrong electronic interactions? If it works, the macroscopic NLOeffect should be further enhanced. To answer this question, P132,constructed by two different chromophore moieties with the reg-ular AB structure (Chart 42) was prepared via Sonogashira couplingreaction. For comparison, P129eP131 (Chart 42) were also syn-thesized. As shown in Chart 42, P132 exhibited the highest d33value (116.8 pm/V), while the d33 value of its analogue, P131, withthe irregular structure, was only 45.1 pm/V (between the values ofP129 and P130). This indicated that the sulfonyl-based chromo-phore could act as effective isolation group for the nitro-based one,the main chromophore moieties, but, only in the case of P132 withthe regular AB structure. On the other hand, due to the presence ofthe isolation chromophore with lower mb value, the maximumabsorption of the P132was 30 nmblue-shifted (in THF solutions), incomparison with P129 just containing one type of chromophore.The blue-shifted maximum absorption would result in the wideoptical transparency window, and benefit to the practicalapplications.
To further investigate the interactions between two types ofchromophores, according to the concept of SIG, we introduceddifferent isolation groups from small size (H) to larger ones(carbazole, Cz) into themain chromophore (nitro-based one) (Chart
N
N
O
Isolation C
N
NN
n
d33
= 34.3 pm/V
SO OC2H5
N
NN
NO2
n
d33
= 78.1 pm/V
N
NP129
P130
Chart 42. The structures and d33 valu
43) [46a]. To our surprise, in this series of polymers, naphthaline(Np)was the SIG, and the d33 value of P135was up to 122.1 pm/V. Asmentioned above, Cz was usually the suitable isolation group fornitro-based azo chromophore (P8 in Table 1 as an example) [18a].Thus, to overcome the strong intramolecular dipoleedipole in-teractions among the chromophores in these NLO polymers con-taining isolation chromophore, much smaller isolation groups weresuitable (Np per two chromophores vs Cz per one chromophore),indicating that the dipoleedipole interactions became muchsmaller by the introduction of the isolation chromophore.Furthermore, the d33 value of P135 was nearly 40 pm/V higher thanthat of P8, further confirming the advantages of isolation chromo-phore. We also applied the concept of “isolation chromophore” tohyperbranched polymers, and achieved success [46b].
Then, we wondered what would happen if the above twochromophores were utilized to construct dendrimers with theregular structure similar to P132, while the formed triazole ringsact as “the suitable isolation group” just as in the case of G1eG5.Perhaps, much better NLO effect could be achieved. Similar toG1-PFPh to G5-PFPh, hyperbranched polymers were also used asmodels. As shown in Chart 44, hyperbranched polymers P137and P138 were synthesized derived from an AB4-type monomercontaining two chromophores with regular structure [47a]. Theresult showed that the main chromophores in the periphery(P138) would enhance its d33 value in a large degree, while theNLO coefficient of the hyperbranched polymers containingisolation chromophores in the periphery exhibited nearly noimprovement (P137). Thus, dendrimers G2-NS to G5-NS (Chart45, NS means the nitro- and sulfonyl-based chromophoreswere regular arranged to differentiate G1eG5) with nitro-basedazo chromophore in the periphery were prepared [47b]. Incomparison with G5, which only contained one kind of chro-mophore (the nitro-based one), G5-NS demonstrated muchhigher NLO coefficient values (253.0 vs 193.1 pm/V). This shouldbe due to the presence of the isolation chromophore (thesulfonyl-based one), similar to the case of P132. It was under-standable. The mb value of the sulfonyl-based chromophore islower than that of the nitro-based one, thus, the sulfonyl-basedmoieties and the triazole groups could act as the isolationgroups to facilitate the noncentrosymmetric alignment of the
N
S OC2H5
N OO
NN
NO2
O On
hromophore
Main Chromophore
d33
= 116.8 pm/V
N
NO2
N
NN
SO OC2H5
0.5 0.5
d33
= 45.1 pm/V
P131
P132
es of NLO polymers P129eP132.
P133-P136
N
NN
NO2
N OO
NN
O On
R
SOC2H5
OMain
Chromophore
Isolation
Chromophore
Changeable
Isolation Group
P134: R=P133: R= HO
OO
P135: R=
O
OO
P136: R=
O
OO
N
Chart 43. The structures of NLO polymers P133eP136 with isolation chromophore.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4379
nitro-based chromophore moieties in the dendrimers under theelectronic field, to contribute to the macroscopic NLO effect. Onthe other hand, the sulfonyl-based chromophore groups them-selves could be well isolated to contribute to the macroscopicNLO effect, by the triazole moieties, which should be big enoughto shield the electronic interactions between the chromophoremoieties due to their relatively low mb value. In addition, due tothe presence of sulfonyl-chromophores, all the maximum ab-sorptions of the dendrimers contain the isolation chromophorewere blue-shifted, in comparison with that only containing one
(CH2)3
NN
N
(CH2)3
N NN
(CH2)3
NN N
(CH2)3
(CH2)3
NNN
(H2C)3
NN
N
(H2C)3
NNN
(CH2)3
NN N
= polymer chain
(CH2)3NN
N
(CH2)3NN
N
(CH2)3NN
N
(H2C)3NN
N
(H2C)3
NNN
(H2C)3
NNN
(H2C)3
NN
N
NN3 N3
NN
NO2
ON
N
NN
SOO
NN
NN3N3
NN
NO2
ONN N
CuSO4
NaASCDMF
End-Capped Chromophore
P127 and P128
O
O
End-Capped
P127 d33
= 117.6 pm/V
P128 d33
= 167.4 pm/V
Chart 44. The structures and d33 values of NLO hyperbr
type of chromophore, and the blue-shifted maximum absorp-tions of dendrimers would result in their wide optical trans-parency window, and benefit to the practical application inphotonics fields. Very recently, we combined the utilization ofthe isolation chomophore and AreArF self-assembly effect, tosynthesize a new series of dendrimers G2-PFPh-NS to G5-PFPh-NS, and the d33 values were further improved. The d33 value ofG5-PFPh-NS was up to 257 pm/V, which should be the new re-cord reported so far for polymers containing simple azo chro-mophore moieties [45b].
NN S
O
O(CH2)3
NN
NO2
O(CH2)3
=
=
NNN=
O
O
O
O
NNN=
O
O
O
O
P127
P128
SO
O(CH2)3
O(CH2)3
NO2
N
NN
SOO
O
O
N
NO2
O
NN N N
NNN
N
NOO
O O
N
NO2
O
N
NO O
OO
NN
SOO
Chromophore for P127 End-Capped Chromophore for P128
anched polymers derived from AB4-type monomer.
OR
N NN
O
OR
NNN
O
OR
N NN
O
OR
NNN
O
NN
N
NN
N
N
O
O O
O
O
O O
O
G2-NS
R
O O
NN N
RN
NN
R O
NNN
R
O O
NN N
RN
NN
R O
O
NNN
R
OO
NNN
RNN
N
RO
N NN
R
OO
NNN
RNN
N
RO
O
N NN
NN
N
NN
N
N
OO
OO
O
O
O
O
OO
OO
O
O
O
O
OO
G3-NS
G4-NS G5-NS
= N NN
ONO2
= N NN S
O
O
R = (CH2)3
Chart 45. The structures of NLO dendrimers G2-NS to G5-NS.
W. Wu et al. / Polymer 54 (2013) 4351e43824380
7. Conclusion
This article reviews some recent progress of second-ordernonlinear optical polymers and dendrimers. Due to the lengthlimitation, we havemainly focused on the design of materials in themolecular level. Some other outstanding works on this NLO fieldhave been cited in the references for interested readers [48]. Forexamples, “click chemistry” reactions (including DielseAldercycloaddition reaction and Huisgen cycloaddition reaction) wereused in the cross-linked systems to achieve ultrahigh EO perfor-mance and stability by Jen group or Odobel group; Liu and Zhenet al. designed a new type of chromophore, which could achieveultrahigh r33 value of 337 pm/V in the simple guest-host dopingsystem; Jen et al. developed a novel process using surface-modifiedlithium niobate and lithium tantalate crystals as an effectiveconformal and detachable electric field source for efficient poling ofEeO polymers and so on.
So far, it is still a big challenge to efficiently translate the large b
values of the organic chromophores into high macroscopic NLOactivities of polymers, due to the strong intermolecular electronicinteractions between the chromophore moieties in the polymericsystem. The site-isolation principle and the concept of “suitableisolation group” should be very useful for the rational design of newNLO polymers. “H-type” chromophores and “star-type” chromo-phores were two typical successful examples by utilization of thesite-isolation principle and the concept of “suitable isolationgroup”. In recent years, dendritic macromolecules, including den-drimers and hyperbranched polymers, were considered as a very
promising molecular topology for the next generation of highlyefficient NLO materials. The latest researches demonstrated thatthe AreArF self-assembly effect and isolation chromophore shouldbe two new avenues to achieve high comprehensive performanceof NLO materials, and by using these methods in dendrimers, larged33 value (257 pm/V) was achieved by using simple azo chromo-phores. Thus, it was reasonable to predict that if these methodswere applied to the NLO systems bearing much more polar chro-mophores (FTC, CLD, etc.), even much better macroscopic NLO ef-fects could be realized.
Acknowledgement
We are grateful to the National Science Foundation of China (no.21034006) for financial support.
References
[1] Franken FA, Hill AE, Peters CW, Weinreich. Phys Rev Lett 1961;7:118.[2] (a) Burland DM, Miller RD, Walsh CA. Chem Rev 1994;90:31;
(b) Marder SR, Perry JW. Science 1994;263:1706;(c) Service R. Science 1995;267:1918.
[3] For example, see: (a) Chen C,WuB, Jiang A, You G. Scientia Sinica B 1985;28:235;(b) Chen C, Wu Y, Jiang A, Wu B, You G, Li R, et al. J Opt Soc Am B 1989;6:616;(c) Smith WL. Appl Opt 1977;16:798;(d) Boyd GD, Buehler E, Storz FG. Appl Phys Lett 1971;18:301;(e) Chemla DS, Kupecek PJ, Robertson DS, Smith RC. Opt Commun 1971;3:29.
[4] Clays K, Coe BJ. Chem Mater 2003;15:642.[5] (a) Meredith GR, Van Dusen JG, Williams DJ. Acs Symp Ser 1983;233:109;
(b) Meredith GR, Van Dusen JG, Williams DJ. Macromolecules 1982;15:1385.
W. Wu et al. / Polymer 54 (2013) 4351e4382 4381
[6] Dalton L, Harper A, Ren A, Wang F, Todorova G, Chen J, et al. Ind Eng Chem Res1999;38:8.
[7] Ye C, Fang S. Polym Bull 1990;3:95 [in Chinese].[8] For some very recent reviews, see: (a) Cho MJ, Choi DH, Sullivan PA,
Akelaitis AJP, Dalton LR. Prog Polym Sci 2008;33:1013;(b) Dalton LR, Sullican PA, Denise HB. Chem Rev 2010;110:25;(c) He GS, Tan LS, Zheng Q, Prasad. Chem Rev 2008;108:1245;(d) Jang SH, Jen AKY. Chem Asian J 2009;4:20;(e) Luo J, Zhou X, Jen AKY. J Mater Chem 2009;19:7410;(f) Li Z, Li Q, Qin J. Polym Chem 2011;2:2723.
[9] (a) Dalton LR, Harper AW, Robinson BH. Proc Natl Acad Sci USA 1997;94:4842;(b) Yu D, Gharavi A, Yu L. J Am Chem Soc 1995;117:11680;(c) Sandhya KY, Pillai CKS, Tsutsumi N. Prog Polym Sci 2004;29:45;(d) Wang F, Harper AW, Lee MS, Dalton LR. Chem Mater 1999;11:2285.
[10] Harper AW, Sun S, Dalton LR, Garner SM, Chen A, Kulluri S, et al. J Opt Soc AmB 1998;15:329.
[11] (a) Robinson BH, Dalton LR. J Phys Chem A 2000;104:4785;(b) Robinson BH, Dalton LR, Harper HW, Ren A, Wang F, Zhang C, et al. ChemPhys 1999;245:35;(c) Dalton LR, Steier WH, Robinson BH, Zhang C, Ren A, Garner S, et al. J MaterChem 1999;9:1905.
[12] Luo J, Ma H, Haller M, Barto RR. Chem Comm 2002;8:888.[13] For example, see: (a) Fréchet JMJ. Proc Natl Acad Sci USA 2002;99:4782;
(b) Fréchet JMJ, Hawker CJ, Gitsov I, Leon JW. J Macromol Sci Part A Pure ApplChem 1996;A33:1399;(c) Ma H, Jen AKY. Adv Mater 2001;13:1201;(d) Shi Z, Luo J, Huang S, Zhou X, Kim TD, Cheng YJ, et al. ChemMater 2008;20:6372;(e) Okuno Y, Yokoyama S, Mashiko S. J Phys Chem B 2001;105:2163;(f) Yokoyama S, Nakahama T, Otomo A, Mashiko S. J Am Chem Soc 2000;122:3174;(g) Campbell VE, In I, McGee DJ, Woodward N, Caruso, Gopalan P. Macro-molecules 2006;39:957.
[14] For example, see: Zhou X, Luo J, Huang S, Kim TD, Shi Z, Cheng Y, et al AdvMater 2009;21:1976.
[15] (a) Luo J, Liu S, Haller M, Liu L, Ma H, Jen AKY. Adv Mater 2002;14:1763;(b) Luo J, Haller M, Ma H, Liu S, Kim TD, Tian T, et al. J Phys Chem B 2004;108:8532.
[16] (a) Shi Z, Hau S, Luo J, Kim TD, Tucker NM, Ka JW, et al. Adv Funct Mater2007;17:2557;(b) Kim TD, Luo J, Ka JW, Hau S, Tian Y, Shi Z, et al. Adv Mater 2006;18:3038.
[17] Moylan CR, Miller RD, Twieg RJ, Lee VY, McComb LH, Ermer S, et al. Soc OptEng 1995;2527:150.
[18] For example, see: (a) Li Z, Li Z, Di C, Zhu Z, Li Q, Zeng Q, et al. Macromolecules2006;39:6951;(b) Li Z, Zeng Q, Li Z, Dong S, Zhu Z, Li Q, et al. Macromolecules 2006;39:8544;(c) Li Q, Li Z, Zeng F, Gong W, Li Z, Zhu Z, et al. J Phys Chem B 2007;111:508;(d) Li Q, Li Z, Ye C, Qin J. J Phys Chem B 2007;112:4928;(e) Li Z, Dong S, Yu G, Li Z, Liu Y, Ye C, et al. Polymer 2007;48:5520;(f) Li Z, Li P, Dong S, Zhu Z, Li Q, Zeng Q, et al. Polymer 2007;48:3650;(g) Zhu Z, Li Q, Zeng Q, Li Z, Li Z, Qin J, et al. Dyes Pigm 2008;78:199;(h) Li Z, Yu G, Dong S, Wu W, Liu Y, Ye C, et al. Polymer 2009;50:2806;(i) Li Z, Wu W, Hu P, Wu X, Yu G, Liu Y, et al. Dyes Pigm 2009;81:264;(j) Li Z, Wu W, Ye C, Qin J, Li Z. J Phys Chem B 2009;113:14943;(k) Li Z, Zeng Q, Yu G, Li Z, Ye C, Liu Y, et al. Macromol Rapid Commun2008;29:136;(l) Zeng Q, Li Z, Li Z, Ye C, Qin J, Tang B. Macromolecules 2007;40:5634.
[19] (a) Ma H, Liu S, Luo J, Suresh S, Kang SH, Haller M, et al. Adv Funct Mater2002;12:565;(b) Hammond SR, Clot O, Firestone KA, Bale DH, Lao D, Haller M, et al. ChemMater 2008;20:3425;(c) Chang P, Chen J, Tsai H, Hsiue. J Polym Sci Part A: Polym Chem 2009;47:4937;(d) Kim TD, Luo J, Tian Y, Ka JW, Tuker NM, Haller M, et al. Macromolecules2006;39:1676.
[20] Li Q, Yu G, Huang J, Liu H, Li Z, Ye C, et al. Macromol Rapid Commun 2008;29:798.[21] (a) Li Z, Hu P, Yu G, Zhang W, Jiang Z, Liu Y, et al. Phys Chem Chem Phys
2009;11:1220;(b) Li Z, Qiu G, Ye C, Qin J, Li Z. Dyes Pigm 2012;94:16.
[22] (a) Döbler M, Weder C, Neuenschwander P, Stuter UW, Follonier S, Bosshard C,et al. Macromolecules 1998;31:6184;(b) Döbler M, Weder C, Ahumada O, Neuenschwander P, Stuter UW,Follonier S, et al. Macromolecules 1998;31:7676.
[23] Li Z, Wu W, Yu G, Liu Y, Ye C, Qin J, et al. ACS Appl Mater Inter 2009;1:856.[24] (a) Zhang CZ, Lu C, Zhu J, Lu G, Wang X, Shi Z, et al. Chem Mater 2006;18:
6091;(b) Zhang CZ, Lu C, Zhu J, Wang C, Lu G, Wang C, et al. Chem Mater 2008;20:4628;(c) Shi J, Wu D, Ding Y, Wu D, Hu H, Lu G. Tetrahedron 2012;68:2770.
[25] (a) Gopalan P, Katz HE, McGee DJ, Erben C, Zielinski T, Bousquet D, et al. J AmChem Soc 2004;126:1741;(b) Campbell VE, In I, McGee DJ, Woodward N, Caruso A, Gopalan P. Macro-molecules 2006;39:597.
[26] (a) Ma H, Chen BQ, Sassa T, Dalton LR, Jen AKY. J Am Chem Soc 2001;123:986;(b) Sullivan PA, Akelaitis AJP, Lee SK, McGrew G, Lee SK, Choi DH, et al. ChemMater 2006;18:344;(c) Pereverzev YV, Prezhdo OV, Dalton LR. ChemPhysChem 2004;5:1821;(d) Sullivan PA, Rommel H, Liao Y, Olbricht BC, Akelaitis AJP, Firestone KA,et al. J Am Chem Soc 2007;129:7523;(e) Pereverzev YV, Gunnerson KN, Prezhdo OV, Sullivan PA, Liao Y,Olbricht BC, et al. J Phys Chem C 2008;112:4355.
[27] (a) Gao J, Cui Y, Yu J, Lin W, Wang Z, Qian G. J Mater Chem 2011;21:3197;(b) Lin W, Cui Y, Gao J, Yu J, Liang T, Qian G. J Mater Chem 2012;22:9202.
[28] Wu W, Wang C, Zhong C, Ye C, Qiu G, Qin J, et al. Polym Chem 2013;4:378.[29] (a) Li Z, Yu G, Wu W, Liu Y, Ye C, Qin J, et al. Macromolecules 2009;42:2864;
(b) Li Z, Wu W, Li Q, Yu G, Xiao L, Liu Y, et al. Angew Chem Int Ed 2010;49:2763.
[30] (a) Wu W, Huang L, Song C, Yu G, Ye C, Liu Y, et al. Chem Sci 2012;3:1256;(b) Wu W, Wang C, Tang R, Fu Y, Ye C, Qin J, et al. J Mater Chem C 2013;1:717.
[31] For some recent reviews, see: (a) Voit BI, Lederer A. Chem Rev 2009;109:5924;(b) Qin A, Lam JWY, Tang B. Chem Soc Rev 2010;39:2522;(c) Mintzer MA, Grinstaff MW. Chem Soc Rev 2011;40:173;(d) Häubler M, Qin A, Tang B. Polymer 2007;48:6181;(e) . Special Issue Branched polymers: Frey H. guest ed Macromol Chem Phys2007;208:1607.
[32] (a) Zhang Y, Wada T, Sasabe H. Polymer 1997;38:2893;(b) Zhang Y, Wang L, Wada T, Sasabe H. Macromol Chem Phys 1996;197:667;(c) Zhang Y, Wang L, Wada T, Sasabe H. J Polym Sci Part A Polym Chem1996;34:1359.
[33] Bai Y, Song N, Gao J, Sun X, Wang X, Yu G, et al. J Am Chem Soc 2005;127:2060.
[34] (a) Li Z, Lam JWY, Dong Y, Dong Y, Tang BZ, Qin A, et al. Polym Pre2004;228:350;(b) Li Z, Qin A, Lam JWY, Dong Y, Dong Y, Ye C, et al. Macromolecules2006;39:1436.
[35] (a) Zhu Z, Li Z, Tan Y, Li Q, Zeng Q, Ye C, et al. Polymer 2006;47:7881;(b) Li Z, Wu W, Ye C, Qin J, Li Z. Macromol Chem Phys 2010;211:916;(c) Li Z, Wu W, Ye C, Qin J, Li Z. Polym Chem 2010;1:78.
[36] Li Z, Wu W, Ye C, Qin J, Li Z. Polymer 2012;53:153.[37] Scheel AJ, Komber H, Voit BI. Macromol Rapid Commun 2004;25:1175.[38] (a) Li Z, Yu G, Hu P, Ye C, Liu Y, Qin J, et al. Macromolecules 2009;42:1589;
(b) Li Z, Wu W, Qin G, Yu G, Liu Y, Ye C, et al. J Polym Sci Part A Polym Chem2011;49:1977;(c) Wu W, Ye C, Qin J, Li Z. Polym Chem 2013;4:2361.
[39] Scarpaci A, Blart E, Montembault V, Fontaine L, Rodriguez V, Odobel F. ACSAppl Mater Inter 2009;1:1799.
[40] Unpublished work.[41] Patrick CR, Prosser GS. Nature 1960;187:1021.[42] Coates GW, Dunn AR, Henling LM, Dougherty DA, Gubbs RH. Angew Chem Int
Ed 1997;36:248.[43] (a) Kim TD, Kang J, Luo J, Jang S, Ka J, Tucker N, et al. J Am Chem Soc
2007;129:488;(b) Gray T, Kim TD, Knorr Jr DB, Luo J, Jen AKY, Overney RM. Nano Lett2008;8:754.
[44] (a) Wu W, Huang Q, Qiu G, Ye C, Qin J, Li Z. J Mater Chem 2012;22:10486;(b) Wu W, Zhu Z, Qiu G, Ye C, Qin J, Li Z. J Polym Sci Part A Polym Chem2012;50:5124.
[45] (a) Wu W, Fu Y, Wang C, Ye C, Qin J, Li Z. Chem Asian J 2011;6:2787;(b) Wu W, Yu G, Liu Y, Ye C, Qin J, Li Z. Chem Eur J 2013;19:690.
[46] (a) Wu W, Ye C, Qin J, Li Z. Chem Asian J 2013. http://dx.doi.org/10.1002/asia.201300010;(b) WuW, Huang L, Xiao L, Huang Q, Tang R, Ye C, et al. RSC Adv 2012;2:6520.
[47] (a) Wu W, Ye C, Yu G, Liu Y, Qin J, Li Z. Chem Eur J 2012;18:4426;(b) Wu W, Li C, Ye C, Yu G, Liu Y, Qin J, et al. Chem Eur J 2012;18:11019.
[48] (a) Cabanetos C, Bentoumi W, Silvestre V, Blart E, Pellegrin Y, Montembault V,et al. Chem Mater 2012;24:1143;(b) Wu J, Bo S, Liu J, Zhou T, Xiao H, Liu T, et al. Chem Comm 2012;48:9637;(c) Huang S, Luo J, Yip HL, Ayazi A, Zhou X, Gould M, et al. Adv Mater 2012;24:OP42;(d) Shi Z, Luo J, Huang S, Polishak BM, Zhou X, Liff S, et al. J Mater Chem2012;22:951;(e) Hammond SR, Sinness J, Dubbury S, Fireston KA, Benedict JB, Wawezak Z,et al. J Mater Chem 2012;22:6752;(f) Zhou X, Luo J, Davies JA, Huang S, Jen AKY. J Mater Chem 2012;22:16390;(g) Huang S, Luo J, Jin Z, Zhou X, Shi Z, Jen AKY. J Mater Chem 2012;22:20353;(h) Zhang Y, Ortega J, Baumeister U, Folcia C, Sanz-Enguita G, Walker C, et al.J Am Soc Chem 2012;134:16298.
er 54
Wenbo Wu is pursuing his Ph.D. in the Department ofChemistry at Wuhan University, under the supervision ofProf. Zhen Li. He received his Bachelor’s degree of Sciencefrom Wuhan University, China in 2009. His research fo-cuses on the organic second-order nonlinear optical mate-rials and the synthesis of hyperbranched polymers anddendrimers.
W. Wu et al. / Polym4382
Jingui Qin obtained his B.Sc. and M.Sc. in Department ofChemistry at WHU. After he obtained D.Phil. degree inOrganometallic Chemistry from University of Oxford in UKunder Prof. Malcolm Green FRS in 1987, he returned toWHU and became a full professor in 1993. He has carriedout collaboration research in the Institute of Physical andChemical Research, Japan, the Institute of Molecular Sci-ences, Japan and California Institute of Technology, USA.His interests are in the design, synthesis, structure andproperties of various organic, inorganic and polymericopto-electronic materials.
Zhen Li received his B.Sc. and Ph.D. degrees from WuhanUniversity (WHU) in China in 1997 and 2002, respectively,under the supervision of Prof. Jingui Qin. In 2003e2004,he worked in the Hongkong University of Science andTechnology as Research Associate in the group of Prof. BenZhong Tang. In 2010, he worked in Georgia Institute ofTechnology in the group of Prof. Seth Marder. He is a fullprofessor at WHU from 2006, and his research interestsare in the development of organic molecules and polymerswith new structure and new functions for organic elec-tronics and photonics.
(2013) 4351e4382
