A Computational Probe into the Dissolution Inhibitation Effect of

Dec 23, 2016 - A. Maceiras , T. Marinho , José Luis Vilas , Enrique Carbó-Argibay , Yury ... Lily Liu , Changbin Song , Bin Xue , Jing Li , Junxi Wa...
0 downloads 0 Views 2MB Size
Subscriber access provided by GAZI UNIV

Article

A Computational Probe into Dissolution Inhibitation Effect of Diazonaphthoquinone Photoactive Compounds on Positive Tone Photosensitive Polyimides Feng Zheng, Cornelia G. C. E van Sittert, and Qinghua Lu J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.6b11316 • Publication Date (Web): 23 Dec 2016 Downloaded from http://pubs.acs.org on January 3, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry C is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

A Computational Probe into Dissolution Inhibitation Effect of Diazonaphthoquinone Photoactive Compounds on Positive Tone Photosensitive Polyimides Feng Zheng†, Cornelia G. C. E. van Sittert‡, Qinghua Lu*† †

School of Chemistry and Chemical Engineering, The State KeyLaboratory of MetalMatrix

Composites, Shanghai Jiaotong University, 800 Dongchuan Road, Minhang District, Shanghai, 200240,P. R. China. ‡

Laboratory for Applied Molecular Modelling, Chemical Resource Beneficiation Focus

Area, North-West University, Private Bag x6001, Potchefstroom, 2520, South Africa.

*Corresponding Author Email address: [email protected]

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 38

Abstract:

Positive tone photosensitive polyimides (p-PSPIs) composed of poly(amic acid) (PAA) and a diazonaphthoquinone photoactive compound (DNQ PAC) have been greatly contributed to the progress of microelectronics. However the relationships among PAC molecular structures, hydrogen bonding interactions and dissolution inhibitation for p-PSPIs have not been well understood. In this study, multiscale molecular modeling was utilized to evaluate such relationships. Density functional theory (DFT) calculations were used to predict the polarity of various DNQ PAC models and their corresponding indenylidene ketene (IK) compounds. Molecular dynamics (MD) simulations were performed to mimic the interactions between DNQ PAC and PAA polymer chains by calculating parameters such as the energy of mixing (∆Emix) and Flory-Huggins parameter (χAB). The computational results showed that χAB values of PACs containing mono functional phenols significantly differed before and after UV exposure. Their corresponding suppositional p-PSPI films were found to form a “skin layer” by covering high concentration of PAC on the surface of the film. Experimental dissolution behavior measurements of selected p-PSPI films strongly supported the computational observations. Succinctly, this work demonstrated the applicability of atomistic molecular simulations for the evaluation of dissolution inhibitation effect of DNQ PACs and to understand the possible dissolution inhibition mechanisms of p-PSPIs.

ACS Paragon Plus Environment

2

Page 3 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

1. Introduction In recent years photosensitive polyimides (PSPIs) have greatly contributed to the progress of microelectronics as packing and insulating materials. In comparison with conventional photoresists, the advantages of PSPIs are not only demonstrated by the outstanding properties of polyimides (PIs), but also by the significant simplification of the pattern formation process.1-2 Although the first commercial PSPIs were of the negative tone type,3-4 the positive tone PSPIs (p-PSPIs) are more desired on industrial level. A variety of p-PSPIs have been developed using different chemistries,5-11 in which the most popular ones are those composed of a PI precursor, poly(amic acid) (PAA) and a diazonaphthoquinone photoactive compound (DNQ PAC). In an effort to design PAA-DNQ type p-PSPIs with high contrast imaging, many approaches have been investigated.12-17 These approaches can be divided into two aspects: modification of the PAA structures12-14 and design new PACs.15-17 No matter which aspect was considered, it is inevitable to understand the interactions between PAA and PACs, and their dissolution inhibitation mechanisms for the design of good p-PSPIs. Based on literature, two possible inhibitation mechanisms have been proposed for p-PSPIs, namely the novolac-type hydrogen bonding mechanism18 and the so called “skin layer” mechanism.19-22 Scheme 1 represents the novolac-type hydrogen bonding mechanism. The formation of a hydrophobic “macromolecular complex” by strong intermolecular hydrogen bonding between the hydroxyl groups of the resist and the DNQ moieties of the PAC could lead to high inhibitation.23 The inhibition ability was found to correlate with the structural features of both the resists and the PACs.24 Applying a similar inhibitation mechanism to PAA-DNQ type p-PSPIs enables us to easily understand the pattern formation process. However, the mechanism must be somewhat different due to the different molecular structures of PAA and the smaller dissolution

ACS Paragon Plus Environment

3

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 38

contrast ratio. No clear correlation has been established between the dissolution inhibitation with chemical structures of PAA and DNQ PACs.

Scheme 1. Novolac-type dissolution inhibitation mechanism and the photolysis of DNQ. On the other hand, the protectable “skin layer” mechanism (Figure 1) was proposed according to some experimental studies on p-PSPI films using DNQ PACs with bisphenol or trisphenol backbones (Figure 2).8,19 Two different dissolution rates between the surface and the bottom layers of these p-PSPI films were observed. This could be due to the fact that a high concentration of PAC is distributed on the film surface after prebake (PB). The PAC-rich surface works as a dissolution-inhibiting protection layer because of the hydrophobic property of the

ACS Paragon Plus Environment

4

Page 5 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

PACs. However, the intrinsic behavior and exact structural changes of the films before and after UV exposure has not been reported yet.

Figure 1. The schematic diagram of the “skin layer” mechanism. R HO

OH R R = CH3 , CF3 Bisphenol OH

OH

OH Trisphenol

Figure 2. Bisphenol and trisphenol backbones for preparing DNQ PACs. In this study, multiscale molecular simulations were utilized to evaluate the relationships among PAC molecular structures, hydrogen bonding interactions and dissolution inhibitation, as well as to understand the intrinsic behavior of the p-PSPI films upon UV irradiation. Density

ACS Paragon Plus Environment

5

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 38

functional theory (DFT) calculations were used to predict the polarity of small model structures of DNQ PACs and binding energies of the formation of hydrogen bonds. Taking into account the impact of the complex PAA polymer chain structures, molecular dynamics (MD) simulations were introduced to estimate the interaction energies of DNQ PACs with PAA polymers and to investigate their miscibility, as well as the morphology of the suppositional p-PSPI films. According to the photolysis mechanism of DNQ shown in Scheme 1, DNQ molecule would lose N2 after UV exposure, and convert to its indenylidene ketene analogue. In order to understand the effect of photochemistry on the intrinsic behavior of p-PSPI films, systematical calculations to mimic the exposed model of the films were also conducted. Based on computational evaluations, an experimental dissolution profile study on the selected p-PSPI films were carried out to give a more comprehensive insight on the possible dissolution inhibition mechanisms.

2. Computational and Experimental Methodology 2.1. Calculation Details Density functional theory (DFT) calculations. Molecules shown in Figure 3 were chosen to represent the variation of DNQ PACs, whereas the notation of DNQ PACs represents three general types according to the used backbones, i.e., 2,3,4-trihydroxybenzophenone (TOB) for type I, tetrahydroxybenzophenones for type II, and 1,1,1-tris(4-hydroxyphenyl)ethan (THPE) and its analogues for type III. Due to the fact that the commercial PAC, produced from the reaction between 2,3,4-trihydroxybenzophenone (TOB) and 1,2-naphthoqunione-2-diazo-5sulfonyl chloride (2,1,5-DNQ), is a mixture of I-1, I-2 and I-3 (Scheme 2),25 we therefore need to take account of the properties of all these individual components to evaluate such commercial PAC.

ACS Paragon Plus Environment

6

Page 7 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

O

OD

O

OD

OH

OD

OH

OD

OD

I-1

OD

I-2 O

I-3 OD

DO

OD

DO

OD

II-1

OD

O

O

D= II-2

O S O OD

H

CH3 OD

OD

OD

H

OD

OD III-1

N2

OD

DO

OD

OD

OH

O

CH3

OD III-2

III-3

Figure 3. Chemical structures of DNQ PACs.

ACS Paragon Plus Environment

7

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 38

Scheme 2. Composition of one typical commercial DNQ PAC. The dipole moment of the optimized PAC structures and binding energies of hydrogenbonded PAA-DNQ structures (Scheme S1) were calculated using the DMol3 DFT code26-28 as implemented in BIOVIC Materials Studio 2016. The PW91 functional29 with the large basis set (DNP 3.5) was used to optimize the configuration of the PACs and the hydrogen-bonded PAADNQ structures. The absence of imaginary frequencies for the vibrational frequencies reflected that the structures were at the global minima. Molecular Dynamics (MD). As listed in Table 1, seven p-PSPI film (PF) formulas with DNQ loading ca. 25 wt% to the solid content of PAA were independently explored. All of the calculations were carried out with the condensed-phase optimized molecular potentials for

ACS Paragon Plus Environment

8

Page 9 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

atomistic simulation studies (COMPASSII) force field.30-31 Research has shown that this force field is highly efficient and accurate for the simultaneous of molecular interactions and properties of both gas and condensed phases of a broad range of molecules and polymers.32 All periodic structures were subjected to geometry optimization employing the Forcite module33-35 (a classical molecular mechanics and dynamics tool) with the Smart algorithm, which is a cascade of the steepest descent, a conjugate gradient, quasi-Newton, and adjusted basis set NewtonRaphson methods. Optimization calculations were run continuously until the points of convergence and minimum energy were reached. Table 1. P-PSPI films of different compositions considered in MD simulations Number of Number of Number of System

DNQ Number of

DNQ PAC

Composition PAA units

chains

loading atoms

molecules

(%)

PF1

PAA+I-mix*

10

3

6

1776

24.8

PF2

PAA+I-1

10

3

5

1809

26.8

PF3

PAA+II-1

10

3

4

1806

27.1

PF4

PAA+II-2

10

3

4

1806

27.1

PF5

PAA+III-1

10

3

5

1879

28.4

PF6

PAA+III-2

10

3

4

1830

25.7

PF7

PAA+III-3

10

3

4

1854

26.2

* I-mix components: I-1 x 1 + I-2 x 4 + I-3 x 1

ACS Paragon Plus Environment

9

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 38

Three kinds of systems were calculated by MD simulation, namely, the pure polymer system with three PAA chains, the PAA-DNQ system with three PAA chains and more or less 25 wt% loading of DNQ PAC as highlighted in Table 1, and the pure DNQ PAC system. It has been reported that the minimum number of atoms for a constructed cell is 1000 in polymer calculations.36 In order to minimize the complexity of the calculation and obtain reasonable accuracy with the available computing resource, the PAA polymer chain was fixed at 10 repeat units and three polymer chains were used in all systems. The p-PSPIs were constructed in a dry environment (absence of water or other solvent molecules) to mimic the final product after processing and prebake (PB). In order to build each p-PSPI model films, the Amorphous cell module, which is based more on molecular dynamics principles, was utilized. Computer-aided construction of the p-PSPI films was atom-based and performed under PB conditions (temperature = 130 oC, intrinsic system pressure = 0.1 MPa) and a density ramped from 0.6 to 1.5 g/cm3. All component molecules were loaded into a cubic lattice under periodic boundary conditions,37 and 10 configurations were generated. The configuration with the lowest energy was chosen in subsequent calculations for equilibration using the Forcite module. Each simulation was run with a time step of 1 fs. The temperature was initially set to 405 K and 100 ps of isothermal-isobaric (NPT) ensemble was performed using the configuration sampled with lowest energy as a starting configuration. This was followed by 100 ps of NPT dynamics at 360 K, and the final configuration obtained was then used as the starting configuration of NPT dynamics at 298 K. The simulations were run until the average volume and average energy remained constant (no significant drift) for a minimum of 500 ps representing the equilibration of the system. Trajectories were saved at 5 ps intervals, and the configurations of the final 50 ps were used for data analysis. The three-step MD simulations at gradually decreased temperature

ACS Paragon Plus Environment

10

Page 11 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

was to mimic an equilibration process of the films after PB. The pressure and temperature in the MD simulations were controlled by the Anderson thermostat38 and the Berendsen method,39 respectively. The interactions of electrostatic terms were calculated by Ewald summation, while the interactions of van der Waals terms were estimated by atom-based summation with the buffer width of 0.5 Å, a spline width of 1 Å, and a cutoff distance of 15.5 Å. Figure 4 represents the samples of converged periodic structure of pure PAA film and p-PSPI film PF1 before and after UV exposure. The other periodic structures are shown in Figures S1-S3. The total interaction energy (ET), kinetic energy (EK), and cohesive energy density (CED) were calculated for each structure. The ET represents the sum of the contributory interparticle interactions including attraction and repulsion potentials among component atoms within a particular system, while the Ek is a measure of matrix internal disorder.40

Figure 4. Images of geometrically optimized constructs at the point of convergence: (a) PAA polymer, (b) PF1 before UV exposure, (c) PF1 after UV exposure. The values of cohesive energy density (CED) can quantitatively depict the intensity of the attractive intermolecular forces (miscibility) among the neighboring molecules by using the Flory-Huggins theory.41 The Flory-Huggins interaction parameter χAB is another useful criterion

ACS Paragon Plus Environment

11

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 38

to predict and understand the miscibility of the binary system including a large molecular weight (Mw) component and a small Mw component, such as polymer-solvent,42 polymer-CO243 and polymer-drug,44 as well as two types of polymers.45 For a binary mixture containing polar components and hydrogen bonding interactions, χAB can be evaluated as following: χ = 

∆





(Eq1)

Where the value of Vmono represents the molar volume of small Mw molecule,41, 46 in this case is the molar volume of DNQ PAC molecule. ∆Emix is the energy of mixing when A and B phases are mixed, which can be estimated by Eqs (2) and (3)47-48 ∆ = Φ (

 

) + Φ  (

CED =

 

) − (

 

)

 

(Eq 2)

(Eq 3)

Where ΦA and ΦB are the volume fractions of components A and B, respectively, and ΦA + ΦB = 1. Therefore, χAB can be calculated as long as the CED values for constituent phases are obtained from MD simulations. 2.2. Experimental Details Materials. Poly(amic acid) (PAA) was purchased from YiDun New Material (SuZhou) Co., Ltd, SuZhou, China. 1,2-Naphthoquinone-2-diazo-5-sulfonyl chloride (1,2,5-DNQ), 2,3,4trihydroxybenzophenone

(TOB),

1,1,1-tris(4-hydroxyphenyl)ethane

(THPE),

20%

tetramethylammonium hydroxide (TMAH) aqueous solution and HPLC grade acetonitrile were purchased from Aldrich. 1-methyl-2-pyrrolidinone (NMP) were purchased from Shanghai Chemical Reagents Corp and was dried by molecular sieve before use. The DNQ PACs I-1 and

ACS Paragon Plus Environment

12

Page 13 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

III-1 were selected to prepare experimentally. These PACs were obtained from 1,2,5-DNQ with TOB and THPE, respectively, in a molar ratio of 1:3, according to a reported method.49 The prepared DNQ PACs were named as I-1_E and III-1_E. The p-PSPI films, PF2_E (PAA+I-1_E) and PF5_E (PAA+III-1_E), were obtained by mixing PAA, I-1_E and III-1_E, respectively with solvent (NMP). A 2.38 wt% tetramethylammonium hydroxide (TMAH) aqueous solution was used as a standard developer. Dissolution rate and other measurements. PAC (I-1_E or III-1_E) was added into the PAA resist in NMP (the total solid content was ca. 25 %). The resist films with a thickness of 6.5 µm were obtained by spin-coating from the solutions on silicon wafers. These films were pre-baked (PB) at 130 oC for 5 minutes, then exposed to a Intelli-Ray 600 shuttered UV floodlight lamp (Uvitron International, Inc., Massachusetts, USA) at 365 nm (i-line). The dissolution rate (nm/sec) of the film was determined from changes in the film thickness after the development with 2.38 wt % TMAH in a time interval of 10 sec and followed by rinsing with water. The experimentally prepared PACs, I-1_E and III-1_E were characterized by high-performance liquid chromatography (HPLC). The PAC purity measurement was conducted on a LC-2010A HT HPLC (Shimadzu Corp., Kyoto, Japan) under the following conditions: Column: XBridge BEH300 reversed phase C18 column, 250 x 4.6 mm i.d.; Eluent: HPLC grade acetonitrile and water (1:1); Flow rate: 1 mL/min; Column temperature: 25oC. The contact angles of the p-PSPI films on wafer were measured by placing drops of distilled water on the prepared films using a microscope goniometer (DataPhysics OCA10, DataPhysics Instrument Gmbh, Filderstadt, Germany) at 25oC. The film thickness was measured with a Filmetrics F20 surface profile (Filmetrics, Inc., San Diego, USA).

ACS Paragon Plus Environment

13

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 38

3. Results and Discussions 3.1. Prediction of Polarity in DNQ PACs and Their Indenylidene ketene (IK) Analogues In the case of p-PSPI development, the dissolution inhibitation and dissolution contrast are closely correlated with the polarity of PAC.8,19 In the current study, we carried out polarity prediction on the DNQ PAC structures containing phenolic backbone and DNQ moieties. To estimate the polarity in the DNQ PACs, the dipole moment of the structures was used, whereas a larger dipole moment corresponds to a larger polarity of a compound. DNQ PAC structures given in Figure 1 (DFT-optimized structures see Figure S4) were subjected to dipole moment (µ, Debye) calculations and the obtained results are summarized in Table 2. Table 2. Calculated dipole moment of DNQ PACs and their indenylidene ketene analogues DNQ PAC

Dipole moment (D)

DNQ IK

Dipole moment (D)

I-1

4.15

I-1

7.23

I-2

2.72

I-2

6.55

I-3

4.30

I-3

3.07

I-mix

2.95

I-mix

6.21

II-1

6.08

II-1

9.36

II-2

5.30

II-2

9.92

III-1

3.67

III-1

8.90

III-2

5.92

III-2

9.16

ACS Paragon Plus Environment

14

Page 15 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

III-3

5.67

III-3

9.30

If we use the component ratio as listed in Scheme 2 to estimate an average dipole moment of the commercial DNQ PAC, it would be ca. 2.95 D. Such a small polarity is promising to ensure a good dissolution inhibition. As the third generation DNQ PAC of novolac resist,

compound

III-1 has the smallest polarity among all the compounds, evidenced by the dipole moment value at 3.67 D. This can provide a good dissolution inhibition as well. For the other studied DNQ PACs, the dipole moments are much higher than the two above mentioned PACs, which are expected to give a low dissolution inhabitation. After converting to indenylidene ketene (IK) compounds (DFT-optimized structures see Figure S5) through the photolysis reaction, the dipole moment of DNQ PACs increased dramatically except I-3 (see Table 2). The IK analogue of I-3 was found to have the lowest dipole moment value at 3.07 D. As a minor component in the mixed DNQ PAC (2.5%, see Scheme 2), its polarity would not contribute much to the entire PAC mixture. The total dipole moment value for I-mix was calculated to be 6.55 D after converting. The large dipole moment of IK compounds suggests that the hydrophobic property of these compounds is reduced after UV exposure. 3.2. PAA Polymer Chain – DNQ PAC Interactions The interactions between polymers and small molecules are difficult to be calculated by DFT methods due to the multiscale structures of the polymer and time-consuming calculations. MD simulations were therefore introduced to better understand the intrinsic behavior of the dry

ACS Paragon Plus Environment

15

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 38

p-PSPI film formulations (PF1 to PF7) as well as facilitate the optimal DNQ PAC selection process by predicting the interactions between PAA polymer chains and DNQ PAC molecules. The total energy (ET), the kinetic energy (EK) and the cohesive energy density (CED) of the systems were computed for all the films (Table 3). Table 3. Calculated energies, CED and χ values of PAA-DNQ mixed systems considered in the present MD simulations ET

EK

CED

∆Emix

3

3

System

χAB (kcal/mol)

(kcal/mol)

(J/cm )

(J/cm )

Before UV exposure PF1

182.4

1624.7

584

-19.2

-5.22

PF2

386.1

1663.1

581

-39.4

-10.7

PF3

686.2

1662.7

621

-43.0

-13.9

PF4

383.8

1664.5

682

-39.2

-13.8

PF5

396.7

1723.8

565

-17.3

-5.4

PF6

276.7

1666.9

576

-12.8

-4.6

PF7

281.9

1688.0

507

-13.6

-5.1

-44.3

-5.6

After UV exposure PF1

62.3

1549.5

584

ACS Paragon Plus Environment

16

Page 17 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

PF2

191.1

1572.8

571

-28.1

-11.7

PF3

439.0

1570.2

660

-46.2

-13.9

PF4

288.6

1569.4

682

-50.7

-15.3

PF5

277.8

1633.4

578

-34.7

-9.9

PF6

15.1

1596.9

589

-42.4

-13.4

PF7

132.0

1619.9

562

-39.8

-12.2

It was showed that PF1 (PAA + I-mix) has the lowest ET with the value of 182.4 kcal/mol in all configured p-PSPI films. This indicates the commercial PAC model gives the most stable formulation to form a thermodynamically robust and well-defined film network with a well-established interatomic interaction within the compounds. On the other hand, when using the fully esterified pure tetra-ester II-1 as PAC, the film PF3 was calculated to have the highest total energy (686.2 kcal/mol). This could be due to the steric hindrance of II-1, resulting in a thermally unstable p-PSPI film. For the other in silico films, moderate ET values around 300 kcal/mol were observed. In addition, the computed EK values for all films have same magnitude, showing the chemical structure of DNQ PACs have negligible effect on intramolecular mobility and disarray of the film network. ∆Emix and χAB values were calculated based on the computed cohesive energy density (CED) values of pure PAA polymers, DNQ PACs, and the mixed polymer-DNQ systems according to Eqs 1-3. Table 3 shows that all computed p-PSPI films have negative ∆Emix and

ACS Paragon Plus Environment

17

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 38

χAB values, illustrating a good intermolecular miscibility and compatibility between DNQ PACs and the polymeric PAA component. This is consistent with the experimental observation that the mixtures of PAA and DNQ based photosensitizers are homogenous dark red resists. It has been suggested for novolac photoresists that DNQ causes extensive hydrogen bonding between the phenolic hydroxyl groups to deprotonate when exposed to base. In the system of DNQ and PAA, there are two or three types of oxygen atoms depending on the structure of backbone (i.e., type 1-3 O atoms in Figure 5), and two types of hydrogen atoms (i.e., type a and b H atoms in Figure 5) would have chances to form possible H-bonds. Hydrogen bonding interactions between DNQ PACs and PAA polymers were therefore studied using IR and solid-state NMR spectroscopies. FTIR analysis (Figure S6) showed the H-bonded hydroxyl stretch in the range of 3500 and 3300 cm-1 as reported,50 and

13

C CPMS spectra (Figure S7)

showed the broad peak around 180 ppm due to H-bonded carboxylic acid carbon.51-53 Even though the experimental results proved the hydrogen bonding interactions between DNQ and PAA, it could however not be able to reveal the types of H-bonds.

ACS Paragon Plus Environment

18

Page 19 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 5. Possible H-bonds could be formed in p-PSPI films (Type 3 H-bonds would be only formed in the case of using I and II as PACs). On the other hand, if we look closely into the computed cell for each film, we found that different types of H-bonds were formed between DNQ moieties and polymeric PAA chains as shown in Figure 6, whereas oxygen atoms involved in H-bonding interactions are highlighted with different color according to the type of H-bond. As illustrated in Scheme 1, Type 1a Hbonds would form between DNQ and the carboxylic group in polymer chain in a novolac inhibitation mechanism. Besides Type 1a H-bonds, our computational observation points out that the H-bonds would also form between the sulfonyl group and PAA (Type 2a and 2b), and between the ketone O atom in PAC’s backbone and PAA (Type 3a). Moreover, the formation of H-bonds between sulfonyl group and PAA (Type 2 H-bonds) were found to be most dominant (Table S1), and the selection of H atom type seems to randomly depend on the conformation of the polymeric PAA chains in the film.

ACS Paragon Plus Environment

19

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 38

Figure 6. Different types of hydrogen bonding formed in computed PF2 formulation before UV exposure (Type 1: orange spots, Type 2: green spots and Type 3*: purple spots). * In the case of using I and II DNQ PACs, the formation of Type 3 H-bonds would be possible. Systematical MD simulations were also performed on the suppositional p-PSPI films after UV exposure, i.e., after DNQ moiety converting to its indenylidene ketene (IK) analogue upon the photochemistry. As shown in Table 3, the converted IK compounds bring down the magnitude of both ET and EK in all film formulations, indicating their stability even upon UV irradiation. This is further supported by the decrease in the values of mixing energy and interaction parameter χAB, indicating a better intermolecular miscibility and compatibility between IK compounds and PAA polymer to form a more stable film network after UV

ACS Paragon Plus Environment

20

Page 21 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

exposure. The hydrogen bonding interactions can also be observed in all the formulations of UV exposed films (see Table S1 and Figure S8). After UV irradiation, the converted PAC molecules are distributed much more even in the polymeric PAA, therefore a general increase in the total number of H-bonds was also observed. A slightly decrease of H-bonding strength is evidenced by the elongated distances for all types of H-bonds compared to those of unexposed films. This indicates the H-bonds would be easily broken when the films are soaked in the development to promote the dissolution rate in the exposed area, while the unexposed area may remain the dissolution inhibitation property. In addition, we also compared interaction energy (Eint) to form Type 1 and Type 2 H-bonds, and their bond strength which correlates to the bond length on molecular level using DFT calculations (see Figure S9, Figure S10 and Table S2). It was found that Eint values of these H-bonds except Type 2b have same magnitude before UV exposure. Type 2b H-bond was calculated to be more energetic favored with Eint value at -1.0 kcal/mol. All H-bond lengths calculated are in 1.66 – 2.14 Å range which are typical for stable H-bonds.45 While after UV exposure, these H-bonds in computed films were found to be weakened evidenced by the increased Eint values and the elongated bond lengths. This agrees with the results of MD simulations on UV exposed p-PSPI films. According to the computational results, Type 2b H-bonds could be formed in p-PSPI films most likely. We could therefore hypothesized that the more type2b H-bonds formed between DNQ and PAA, the less hydrophilic channels in the polymers would be shielded, resulting in a poor dissolution inhibition effect. This could be another reason to explain the high solubility of unexposed p-PSPI films composed by PAA and DNQ PACs in aqueous base development besides the low Pka of carboxyl groups in PAA.54

ACS Paragon Plus Environment

21

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 38

3.3. The Dispersivity of DNQ PACs Before and After UV Exposure Hydrophobicity, molecular size and dispersivity of DNQ PACs are reported to be the key factors to improve the dissolution inhibition of a photoresist.55 For p-PSPI films, the dispersivity of DNQ PACs should have a crucial effect besides the H-bonding interactions. We therefore compared the dispersivity of DNQ PACs in the simulated films before and after UV exposure. The simulated morphologies of all unexposed film variants are shown in Figure 7. PF2, PF3 and PF4 represent the films containing fully esterified tri-ester and tetra-esters with benzophenone backbones, I-1, II-1 and II-2, respectively. It clearly shows that DNQ PACs distributed evenly in the entire cell for these film formulations, which can be related to their large negative χAB values (-10.7, -13.9 and -13.8). On the other hand, the trisphenol based DNQ PACs in PF5, PF6 and PF7 concentrated on the top part of the cell, while the bottom part of the cell were only occupied by PAA. These computed morphologies provide a “visible picture” of the so called “skin layer” formation. Due to the fact that the calculated dipole moment of III-2 and III-3 are quite close to those of II-1 and II-2, the difference in the PAC dispersivity between films PF6, PF7 and films PF3, PF4 could mainly be attributed to the chemical structure and size of the PAC molecule. The “skin layer” morphology was also observed in the film formulation PF1 containing I-mix, a model for commercial DNQ PAC, which differs significantly from that of PF2. In this case, the molecular size could not explain the skin layer formation as these esters are analogues to I-1. Therefore, for PF1 and PF5, which contains relatively small size PACs, the top layer formation could be due to their hydrophobic property as supported by their small values of dipole moment (2.90 D and 3.76 D).

ACS Paragon Plus Environment

22

Page 23 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 7. Distribution of DNQ PACs in polymeric PAA matrix before UV exposure. Structures are represented by Corey-Pauling-Koltun (CPK) models, purple for DNQ PACs and green for PAA. When DNQ PACs are converted to their indenylidene ketene (IK) analogues by UV light, the dipole moment of these IK compounds dramatically increased as shown in Table 2, resulting in a low hydrophobic property. Figure 8 illustrates that the in silico morphologies of all UV exposed film variants. They showed the same trends, i.e., PACs are evenly distributed in polymer matrix in all cases except PF1, which was composed with I-mix as PAC. The improved

ACS Paragon Plus Environment

23

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 38

miscibility and compatibility is further supported by the decreased χAB values. In the case of PF1, the smallest dipole moment of I-mix (6.21 D) among the PACs investigated could explain the distribution of the PAC molecules in the film and agrees well with its almost unchanged χAB value before and after UV exposure (-5.22 and -5.6 D), therefore the dispersivity of the PAC varied quite little.

Figure 8. Distribution of DNQs in polymeric PAA matrix after UV exposure. Structures are represented by Corey-Pauling-Koltun (CPK) models, orange for indenylidene ketenes and green for PAA. 3.4. Dissolution Behavior of PAA – DNQ Based p-PSPIs

ACS Paragon Plus Environment

24

Page 25 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

In the current study, a comparison of dissolution behavior was performed on p-PSPI films containing I-1 and III-1, respectively, whereas I-1 represents the 1st generation DNQ PAC containing multiple functional phenol, and III-1 represents the 3rd generation DNQ PAC having three mono functional phenols. Based on the computational results, these two PACs would give different morphologies of PSPI films and then give different profiles on their dissolution behavior. We therefore carried out the measurements on contact angle and dissolution rate of their corresponding p-PSPI films to clarify the simulated observation. After PB at 130 oC for 5 min, the contact angles of the films consisting of PAA, PAA and I1_E or III-1_E (25 wt% to the solid content of polymers) were measured in both unexposed and exposed areas. To be consistent with the in silico films, the experimental films were named as PF2_E (PAA+I-1_E) and PF5_E (PAA+III-1_E), respectively. As shown in Figure 9, the contact angles of the p-PSPI films are higher than that of the PAA film, in which PF5_E has the larger contact angle compared to that of PF2_E before UV exposure. This indicates the surface of resist PF5_E has a higher hydrophobic property due to covering with high concentration of the DNQ PAC. While the distribution of I-1_E in PF2_E should be much more even based on the simulated morphology of PF2, which resulted in a lower hydrophobic surface and a smaller contact angle. After UV exposure, the polarity of PACs are changed, causing the distribution of DNQ PAC molecules to be more homogenetic through the film. Theoretically, the reduced hydrophobicity of the film surface would give a smaller contact angle compared to that of the unexposed area. This is supported by the decreased values of experimentally measured contact angles for the films exposed by UV light as shown in Figure 9.

ACS Paragon Plus Environment

25

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 38

Figure 9. Contact angles of the films after pre-bake (PB). The films such as PAA+I-1_E, PAA +III-1_E were used before (B) and after (A) UV exposure. The PB temperature and time were fixed at 130 oC and 2 min, respectively. Next, the dissolution rate of the films in the developer was measured to investigate the phase separation of the polymer film, and the data are shown in Table 4 and Figure 10. It clearly shows that the film PF5_E has two different dissolution slopes between the surface and the bottom when they was unexposed (Figure 10b), while the exposed film dissolves uniformly. Table 4. Dissolution rates of p-PSPI films PF2_E and PF5_E Dissolution rate (nm/s) Unexposed area

Exposed area

Surface layer

Mid layer

PF2_E

22.7

36.4

68.5

PF5_E

6.1

46.2

66.1

ACS Paragon Plus Environment

26

Page 27 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 10. Dissolution profiles for unexposed and exposed p-PSPI films. The thickness of the surface layer with a very small dissolution rate for PF5_E is about 1000nm, and the rest of the film dissolves at a much larger rate, which is close to that of the exposed ones. It suggests a higher concentration of PAC was distributed on the surface and works as a dissolution-inhibiting layer, coinciding well with our computational observation. In the case of film PF2_E, we also observed two different dissolution rates in the unexposed area, while in silico film PF2 shows a uniform distribution of the PAC. This contradiction can be explained by the composition of experimentally prepared PAC I-1_E, in which there is a mixture of

80%

ACS Paragon Plus Environment

27

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 38

tri-ester I-1 and 20% di-esters (The HPLC charts presented in Figure S7 shows that both PACs are mainly composed of fully esterified product at a ratio of 80.7% for I-1_E, and 91.2% for III-1_E, respectively.). In addition, the difference between the middle layer rate and the surface layer rate is much smaller than that of PF5_E, indicating there should also be some PAC molecules forming H-bond interactions with PAA polymer and inhibiting the dissolution to some extent in the middle layer of the PF2_E. Therefore, the distribution of I-1_E in PF2_E should be the intermediate state of the simulated films PF1 and PF2 (see Figure 7). It should be pointed out that the pure fully esterified compounds for tri- or tetra-hydroxybenzophenone were reported as undesirable DNQ PACs due to their low solubility in common casting solvents,56 therefore the current commercial or lab prepared DNQ PACs tend to be a mixture of esters of such backbones. In other words, the skin layer formation should be always observed by the use of mixed esters as DNQ PAC in experimental site. 3.5. Possible Dissolution Mechanisms The p-PSPIs composed of PAA and DNQ PAC are quite commonly used, it is however difficult to obtain a good pattern due to that the carboxylic acid in PAA chemical structure leads to extremely high solubility in common alkaline developer. If there is no other protection mechanism than the H-bonding inhibition mechanism for such p-PSPIs, the presence of PAC in the PAA polymer would have very minor inhibitation effect to the film in the unexposed area (see Schemes S2&S3). Even though each DNQ moiety in PACs bind with one carboxylic group of PAA, there would still be a free carboxylic group in each PAA repeat unit. The free carboxyl groups in PAA should be still enough to cause the solubility of polymer matrix. This could lead the PAA matrix in p-PSPI films to dissolve in a similar manner as the pure PAA film, resulting in a low dissolution contrast between exposed and unexposed films.

ACS Paragon Plus Environment

28

Page 29 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

With the skin layer protection mechanism, the surface layer become more hydrophobic by covering with high concentration of DNQ PAC molecules and the solubility of this layer would be reduced. On the other hand, the surface would also become more compact due to the strong H-bonding interactions between PAC and PAA polymer. This would prevent or reduce the permeation of the base developer and water through the surface layer, and subsequently reduce the dissolution rate of the unexposed area. After UV exposure, the PACs would distribute uniformly through the film based on computational observation, and promote the dissolution. Above all, the MD simulation results shown that it is better to use the PACs containing several mono functional phenol groups, which should have low to moderate dipole moment and large molecular size. The in silico evaluation of new PACs for p-PSPIs is therefore an important and efficiently way in the designing of p-PSPIs, which can be quickly applied in prior to the experimental work.

4. Conclusion This study investigated the construction, screening, optimization, and evaluation of unexposed and exposed p-PSPI films containing PAA polymer and various DNQ PACs employing both atomistic molecular simulations and experimental methods. Computed results clearly indicated that the chemical structure of PAC has an important effect on the dissolution inhabitation

of

p-PSPI films, in which both the polarity and molecular size play the role. Among the PACs studied, the ones containing backbones with several mono-functional phenols, III-1 to III-3, were calculated to have better dissolution inhibition by the formation of the “skin layer” with gathering of PAC molecules on the film surface. The experimental results on dissolution behavior were well related to the computed outputs. Based on our investigation, the possible

ACS Paragon Plus Environment

29

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 38

dissolution inhibitation mechanism for p-PSPIs is more likely to be an integration of H-bonding interactions and the formation of “skin layer”. In this integrated mechanism, the “skin layer” attributes a dominance protection against the film dissolution in base developer. Furthermore, the “skin layer” is stabilized by the strong H-bonding interactions between DNQ PACs and PAA polymers to form a compact network. Overall, the capability of high performance atomistic molecular modeling approach to provide in-depth understanding of the interactions between PACs and PAA polymers and the possible mechanisms, as well as the screening of PACs for pPSPIs were well established.

Supporting Information. The Supporting Information is available free of charge on the ACS Publications website. Images of geometrically optimized constructs at the point of convergence for PF2 to PF7; Images of geometrically optimized structures of DNQ PACs and indenylidene ketenes; FT-IR and 13C CPMAS spectra for PAA and PF5_E films before and after UV exposure; Table of different hydrogen-bond number in each computed p-PSPI cell; Images of geometrically optimized structures of hydrogen bonded PAA+DNQ structures; HPLC spectrum for PACs I-1_E and III-1_E. ACKNOWLEDGMENT The authors are grateful for financial support from Shanghai Academy of Spaceflight Technology, National program on Key Basic Research Program of China (973 Program: 2014CB643605), and Basic Research Project of Shanghai Science and Technology Commission

ACS Paragon Plus Environment

30

Page 31 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(16JC14039000). The Authors acknowledge Shanghai Jiaotong University for support of this research, and South African Centre for High Performance Computing (CHPC) for donating cluster facility which was used to perform computational work presented in this article.

REFERENCES (1)

Ghosh, M. K. Polyimides: Fundamentals and Applications. Marcel Dekker, Inc.: New

York, 1996. (2)

Photosensitive Polyimides: Fundamentals and Applications; Horie, K.; Yamashita, T.,

Eds.; Technomic Publishing Company, Inc.: Pennsylvania, USA, 1995. (3)

Rubner, R.; Ahne, H.; Kuehn, E.; Kolodziej, G. A Photopolymer - the Direct Way to

Polyimide Patterns. Photographic Sci. Eng. 1979, 23, 303-309. (4)

Yoda, N.; Hiramoto, H. New Photosensitive High Temperature Polymers for Electronic

Applications. J. Macromol. Sci. Chem. 1984, 21, 1641-1663. (5)

Kubota, S.; Moriwaki , T.; Ando, T.; Fukami, A. Positive Photoreactive Polyimides. II.

Preparation and Characterization of Polyimide Precursors Containing α-(2-nitrophenyl)ethyl Ester Side Chains. J. Macromol. Sci. Chem. 1987, 24, 1407-1422. (6)

Omote, T.; Mochizuki, H.; Koseki, K. i.; Yamaoka, T. Fluorine-Containing Photoreactive

Polyimide. 7. Photochemical Reaction of Pendant 1,2-Naphthoquinone Diazide Moieties in Novel Photoreactive Polyimides. Macromolecules 1990, 23, 4796-4802. (7)

Hayase, S.; Takano, K.; Mikogami, Y.; Nakano, Y. High-Yemperature-Post-Exposure

Bake Process (HIT-PEB) for Base-Developable Polyimides Consisting of Diazonaphthoquinones and Polyamic Acids. J. Electrochem. Soc. 1991, 138, 3625-3629. (8)

Hsu, S. L.-C.; Chem, H.-T.; Tsai, S.-J. Novel Positive-Working and Aqueous-Base-

Developable Photosensitive Poly(imide benzoxazole) Precursor. J. Polym. Sci., Polym. Chem.

ACS Paragon Plus Environment

31

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 38

2004, 42, 5990-5998. (9)

Mochizuki, A.; Yamada, K.; Teranishi, T.; Matsushita, K.; Ueda, M. Positive-Working

Photosensitive Polyimide Precursor Based on Polyisoimide Using Nifedipine as a Dissolution Inhibitor. High Perform. Polym. 1994, 6, 225-233. (10)

Fukushima, T.; Oyama, T.; Iijima, T.; Tomoi, M.; Itatani, H. New Concept of Positive

Photosensitive Polyimide: Reaction Development Patterning (RDP). J. Polym. Sci., Polym. Chem. 2001, 39, 3451-3463. (11)

Ogura, T.; Higashihara, T.; Ueda, M. Pattern Formation of Polyimide by Using

Photosensitive Polybenzoxazole as a Top Layer. Eur. Polym. J. 2010, 46, 1576-1581. (12)

Tomikawa, M.; Yoshida, S.; Okamoto, N. Novel Partial Esterification Reaction in

Poly(amic acid) and Its Application for Positive-tone Photosensitive Polyimide Precursor. Polym. J. 2009, 41, 604-608. (13)

Sakayori, K.; Shibasaki, Y.; Ueda, M. A Positive-Type Alkaline-Developable

Photosensitive Polyimide Based on the Polyamic Acid From 2,2',6,6'-Biphenyltetracarboxylic Dianhydride and 1,3-Bis(4-aminophenoxy)benzene, and a Diazonaphthoquinone. Polym. J. 2006, 38, 1189-1193. (14)

Miyagawa, T.; Fukushima, T.; Oyama, T.; Iijima, T.; Tomoi, M. Photosensitive

Fluorinated Polyimides with a Low Dielectric Constant Based on Reaction Development Patterning. J. Polym. Sci., Polym. Chem. 2003, 41, 861-871. (15)

Jha, S.; Reiser, A.; Šroubková, L.; Zahradník, R. A New Design Principle for the

Photoactive Components (PACs) of Dissolution Inhibition Resists. Macromolecules 1998, 31, 6549-6553. (16)

Kawata, S.; Kashiwagi, M.; Mitsuhira, Y.; Kamiya, S. Effective PAC for High

ACS Paragon Plus Environment

32

Page 33 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Performance i-Line Positive Photoresist. J. Photopolym. Sci. Technol. 1991, 4, 101-108. (17)

Koshiba, M.; Murata, M.; Harita, Y. Dissolution Inhibition Mechanisms of

Naphthoquinone Diazides. Polym. Eng. Sci. 1989, 29, 916-919. (18)

Koshiba, M.; Murata, M.; Matsui, M.; Harita, Y. Thermally Induced and Base Catalyzed

Reactions of Naphthoquinone Diazides. SPIE Adv. Resist Technol. Process. 1988, 0920, 364-371. (19)

Tomikawa, M.; Yuba, T.; Ohbayashi, G. Effect of Photoactive Compound Structure on

Photosensitivity of Positive Photosensitive Polyimide. High Perform. Polym. 2006, 18, 603-615. (20)

Inoue, Y.; Higashihara, T.; Ueda, M. Alkaline-Developable Positive-Type Photosensitive

Polyimide Based on Fluorinated Poly(amic acid) and Fluorinated Diazonaphthoquinone. J. Photopolym. Sci. Technol. 2013, 26, 351-356. (21)

Inoue, Y.; Saito, Y.; Higashihara, T.; Ueda, M. Facile Formulation of Alkaline-

Developable Positive-Type Photosensitive Polyimide Based on Fluorinated Poly(amic acid), Poly(amic acid), and Fluorinated Diazonaphthoquinone. J. Mater. Chem. C 2013, 1, 2553-2560. (22)

Inoue, Y.; Ishida, Y.; Higashihara, T.; Kameyama, A.; Ando, S.; Ueda, M. Alkaline-

Developable and Positive-Type Photosensitive Polyimide Based on Fluorinated Poly(amic acid) from Diamine with High Hydrophobicity and Fluorinated Diazonaphtoquinone. J. Photopolym. Sci. Technol. 2014, 27, 211-217. (23)

Shih, H.-Y.; Yeh, T.-F.; Reiser, A.; Dammel, R. R.; Merrem, H. J.; Pawlowski, G. A

Percolation View of Novolak Dissolution. 3. Dissolution Inhibition. Macromolecules 1994, 27, 3330-3336. (24)

Honda, K.; Beauchemin, B. T.; Fitzgerald, E. A.; Jeffries, A. T.; Tadros , S. P.; Blakeney,

A. J.; Hurditch, R. J. Studies of Dissolution Inhibition Mechanism of DiazonaphthoquinoneNovolak Resist II. Effect of Extended Ortho-Ortho Bond in Novolak. SPIE Adv. Resist Technol.

ACS Paragon Plus Environment

33

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 38

Process. 1991, 1466, 141-148. (25)

Fukushima, T.; Hosokawa, K.; Oyama, T.; Iijima, T.; Tomoi, M.; Itatani, H. Synthesis and

Positive-Imaging Photosensitivity of Soluble Polyimides Having Pendant Carboxyl Groups. J. Polym. Sci., Polym. Chem. 2001, 39, 934-946. (26)

Delley, B. An All-Electron Numerical Method for Solving the Local Density Functional

for Polyatomic Molecules. J. Chem. Phys. 1990, 92, 508-517. (27)

Delley, B. Fast Calculation of Electrostatics in Crystals and Large Molecules. J. Phys.

Chem. 1996, 100, 6107-6110. (28)

Delley, B. From Molecules to Solids with the DMol3 Approach. Journal of Chemical

Physics 2000, 113, 7756-7764. (29)

Perdew, J. P.; Wang, Y. Accurate and Simple Analytic Representation of the Electron-Gas

Correlation Energy. Phys. Rev. B,Condens. Matter. 1992, 45, 13244-13249. (30)

Sun, H. COMPASS:  An ab Initio Force-Field Optimized for Condensed-Phase

ApplicationsOverview with Details on Alkane and Benzene Compounds. J. Phys. Chem. B 1998, 102, 7338-7364. (31)

McQuaid, M., J.; Sun, H.; Rigby, D. Development and Validation of COMPASS Force

Field Parameters for Molecules with Aliphatic Azide Chains. J. Comput. Chem. 2004, 25, 61-71. (32)

Fariba, R.; Sepideh, A.-I.; Hamid, M. Molecular Dynamics Simulation Study of Chitosan

and Gemcitabine as a Drug Delivery System. J. Mol. Model. 2015, 21, 165-179. (33)

Hagler, A. T.; Ewig, C. S. On the Use of Quantum Energy Surfaces in the Derivation of

Molecular Force Fields. Comput. Phys. Commun. 1994, 84, 131-155. (34)

Hagler, A. T.; Lifson, S.; Dauber, P. Consistent Force Field Studies of Intermolecular

Forces in Hydrogen-Bonded Crystals.

2.

A Benchmark for the Objective Comparison of

ACS Paragon Plus Environment

34

Page 35 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Alternative Force Fields. J. Am. Chem. Soc. 1979, 101, 5122-5130. (35)

Hagler, A. T.; Stern, P. S.; Sharon, R.; Becker, J. M.; Naider, F. Computer Simulation of

the Conformational Properties of Oligopeptides.

Comparison of Theoretical Methods and

Analysis of Experimental Results. J. Am. Chem. Soc. 1979, 101, 6842-6852. (36)

Belmares, M.; Blanco, M.; Goddard, W. A.; Ross, R. B.; Caldwel, l. G.; Chou, S.-H.;

Pham, J.; Olofson, P. M.; Thomas, C. Hildebrand and Hansen Solubility Parameters from Molecular Dynamics with Applications to Electronic Nose Polymer Sensors. J. Comput. Chem. 2004, 25, 1814-1826. (37)

Theodorou, D. N.; Suter, U. W. Atomistic Modeling of Mechanical Properties of

Polymeric Glasses. Macromolecules 1986, 19, 139-154. (38)

Andersen, H. C., Molecular Dynamics Simulations at Constant Pressure and/or

Temperature. J. Chem. Physics. 1980, 72, 2384-2393. (39)

Berendsen, H. J. C.; Postma, J. P. M.; van Gunsteren, W. F.; DiNola, A.; Haak, J. R.

Molecular Dynamics with Coupling to an External Bath. J. Chem. Physics. 1984, 81, 3684-3690. (40)

Saunders, S. R.; Anand, M.; You, S.-S.; Roberts, C. B. Total Interaction Energy Model to

Predict Nanoparticle Dispersability in CO2-Expanded Solvents. Comput.-Aided Chem. Eng. 2010, 28, 1651-1656. (41)

Flory, P. Thermodynamics of High-Polymer Solutions. J. Chem. Physics. 1942, 10, 51-

61. (42)

Caddeo, C.; Mattoni, A. Atomistic Investigation of the Solubility of 3-Alkylthiophene

Polymers in Tetrahydrofuran Solvent. Macromolecules 2013, 46, 8003-8008. (43)

Hu, D.; Sun, S.; Yuan, P.; Zhao, L.; Liu, T. Evaluation of CO2-Philicity of Poly(vinyl

acetate) and Poly(vinyl acetate-alt-maleate) Copolymers through Molecular Modeling and

ACS Paragon Plus Environment

35

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 36 of 38

Dissolution Behavior Measurement. J. Phys. Chem. B 2015, 119, 3194-3204. (44)

Huynh, L.; Grant, J.; Leroux, J.-C.; Delmas, P.; Allen, C. Predicting the Solubility of the

Anti-Cancer Agent Docetaxel in Small Molecule Excipients Using Computational Methods. Pharm. Res. 2008, 25, 147-157. (45)

Sami, S.; Yildirim, E.; Yurtsever, M.; Yurtsever, E.; Yilgor, E.; Yilgor, I.; Wilkes, G. L.

Understanding the Influence of Hydrogen Bonding and Diisocyanate Symmetry on the Morphology and Properties of Segmented Polyurethanes and Polyureas: Computational and Experimental Study. Polymer 2014, 55, 4563-4576. (46)

Kim, S.-W.; Shi, Y.-Z.; Kim, J.-Y.; Park, K.-N.; Cheng, J.-X. Overcoming the Barriers in

Micellar Drug Delivery: Loading Efficiency, in Vivo Stability, and Micelle-Cell Interaction. Expert Opin. Drug Deliv. 2010, 7, 49-62. (47)

Luo, Z.-L.; Jiang, J.-W. Molecular Dynamics and Dissipative Particle Dynamics

Simulations for the Miscibility of Poly(ethylene oxide)/Poly(vinyl chloride) Blends. Polymer 2010, 51, 291-299. (48)

Jawalkar, S. S.; Raju, K. V. S. N.; Halligudi, S. B.; Sairam, M.; Aminabhavi, T. M.

Molecular Modeling Simulations to Predict Compatibility of Poly(vinyl alcohol) and Chitosan Blends:  A Comparison with Experiments. J. Phys. Chem. B 2007, 111, 2431-2439. (49)

Yuba, T.; Okuda, R.; Tomikawa, M.; Kim, J. H. Soft Baking Effect on Lithographic

Performance by Positive Tone Photosensitive Polyimide. J. Photopolym. Sci. Technol. 2010, 23, 775-779. (50)

Mueller, B. K.; Elce, E.; Grillo, A. M.; Kohl, P. A. Positive-Tone, Aqueous-Developable,

Polynorbornene Dielectric: Lithographic, and Dissolution Properties. J. Appl. Polym. Sci. 2013, 127, 4653-4661.

ACS Paragon Plus Environment

36

Page 37 of 38

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(51)

Miyoshi, T.; Takegoshi, K.; Hikichi, K. High-Resolution Solid State 13C NMR Study of

the Interpolymer Interaction, Morphology and Chain Dynamics of the Poly(acrylic acid)/Poly(ethylene oxide) Complex. Polymer 1997, 38, 2315-2320. (52)

Miyoshi, T.; Takegoshi, K.; Terao, T. Dynamic Alternation between Inter- and Intra-

Polymer Hydrogen Bonds in a Polymer Complex As Studied by Solid-State

13

C 2D Exchange

NMR. Macromolecules 1999, 32, 8914-8917. (53)

Yuan, X.; Xiang, T.-X.; Anderson, B. D.; Munson, E. J. Hydrogen Bonding Interactions

in Amorphous Indomethacin and Its Amorphous Solid Dispersions with Poly(vinylpyrrolidone) and Poly(vinylpyrrolidone-co-vinyl acetate) Studied Using

13

C Solid-State NMR. Mol. Pharm.

2015, 12, 4518-4528. (54)

Hasegawa, M.; Tominaga, A. Environmentally Friendly Positive- and Negative-Tone

Photo-Patterning Systems of Low-K and Low-CTE Polyimides. J. Photopolym. Sci. Technol. 2005, 18, 307-312. (55)

Aoai, T.; Yamanaka, T.; Yagihara, M. Structural Effect of Polymer and Inhibitor on Alkali

Dissolution and Dissolution Inhibition Characteristics. J. Photopolym. Sci. Technol. 1997, 10, 387-396. (56)

DiCarlo, J.; Evans, O. B.; Fedyk, J.; Ficner, S.; Khadim, M.; Dammel, R. R.

Identification of Diazonaphthoquinone Esters of Polyhydroxybenzophenone Compounds. SPIE Adv. Resist Technol. Process. 1994, 2195, 696-706.

ACS Paragon Plus Environment

37

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 38 of 38

Table of Contents Graphic

ACS Paragon Plus Environment

38