Oxidative chemical vapor deposition of polyacenaphthylene, polyacenaphthene, and polyindane via...

Thin Solid Films 556 (2014) 23–27

Contents lists available at ScienceDirect

Thin Solid Films

j ourna l homepage: www.e lsev ie r .com/ locate / ts f

Oxidative chemical vapor deposition of polyacenaphthylene,polyacenaphthene, and polyindane via benzoyl peroxide

Jay J. SenkevichMassachusetts Institute of Technology, Institute for Soldier Nanotechnologies, Bldg NE-47, 5th Floor, 77 Massachusetts Ave, Cambridge, MA 02139, USA

E-mail address: senkej@gmail.com.

0040-6090/$ – see front matter © 2014 Elsevier B.V. All rihttp://dx.doi.org/10.1016/j.tsf.2013.12.051

a b s t r a c t

Article history:Received 22 May 2013Received in revised form 18 December 2013Accepted 19 December 2013Available online 28 December 2013

Keywords:Oxidative CVDOxidative chemical polymerizationAcenaphthyleneAcenaphtheneIndaneThin film

A method is presented to polymerize inexpensive and readily available electron-rich monomers via oxidativechemical vapor deposition. The process uses benzoyl peroxide, an organic oxidant, flash evaporated from achloroform solution via direct liquid injection. The deposition temperatures ranged from 35 °C for indane to125 °C for acenaphthylene and acenaphthene. These temperatures were determined by the volatility andmelting point of the monomers, which were vaporized without the addition of a carrier gas. Benzoyl peroxidewas not cracked into its respective free radical species but instead used as an oxidant at the substratetemperature. Polyacenaphthylene was deposited as a yellow film indicative of its highly conjugated polymerbackbone, whereas polyacenaphthene and polyindane were both transparent dielectrics. Polyacenaphthyleneand polyacenaphthene had higher average indices of refraction 1.6819 and 1.6640 (at 632.8 nm) thanpolyindane 1.5690, likely representing their higher density.

© 2014 Elsevier B.V. All rights reserved.

1. Introduction

Chemical vapor deposited (CVD) polymers have quite a few advan-tages over solution processed polymers such as their ability to be depos-ited as molecular layers [1], deposited without the use of solvents [2],superior optical quality when deposited as thin films, conformality [3],and their general suitability towardmanufacturing processes. Althoughmany polymers can be deposited via the use of plasmas (plasma-en-hanced CVD) [4], their structure often suffers in these processes froma cross-linked structure giving rise to dangling bonds. These danglingbonds in turnmay give rise to poor UV and oxidative stability comparedto the analogous linear chain polymer [5]. Furthermore, thermal CVDmethods have the advantage of preserving functionality in the polymersuch that they can be used in, for example, “click” chemical reactions [6].

Thermal CVD polymers are generally deposited via a free radicaladdition polymerization. Of these polymers, parylene is the mostcommon, where its monomer also acts as the initiator [7]. This initiatorcan be generated in situ via the high temperature, ~650 °C, “cracking” of[2.2]paracyclophane or functionalized derivatives thereof, or withleaving groups that generate the same p-xylylene intermediate [8].Aside from the parylene polymers, thermal CVD polymers have beendeposited via initiated-CVD (i-CVD) and hot-filament CVD (HF-CVD).i-CVD is where a free radical initiator is generated in situ over a filamentarray very near the substrate surface. This free radical initiator initiatespolymerization very effectively with acrylate monomers or monomers

ghts reserved.

that possess double bonds that are very electron poor (have adjacentelectron withdrawing groups) [2]. Unfortunately, this method doesnot work with electron rich monomers. HF-CVD is equivalent to theparylene method of generating the free radical in situ, but it does thisvia a filament array rather than a remote pyrolysis tube in the case ofparylene [9]. This method has been commercialized to depositpoly(tetrafluoroethylene).

Oxidative CVD has been undertaken with inorganic oxidants (FeCl3and Br2) with the focus of polymerizing 3,4 ethylenedioxythiopheneto produce its corresponding polymer, which is the preferentialconducting polymer for organic electronics [2,10]. Unfortunately, bothferric chloride and molecular bromine are very aggressive towardvacuum systems, printed-circuit boards, and biomedical devices. Inaddition, the above methods for depositing polymers yield polymerswith either poor UV and oxidative stability or their chemical processcost is high, for example, parylene AF-4 [11].

Ideally, a method would be developed to deposit thermal CVDpolymers with inexpensive readily available chemistries (monomersor precursors), with a high manufacturing throughput, and uponpolymerization yields polymers with a high UV and oxidative stability.If this process also retained functionality in the polymer chains, itwould be a very powerful technique. The monomers in the currentstudy, acenaphthylene (A), acenaphthene (B), and indane (C), havebeen polymerized via unique methods, including the following:electropolymerization[12], free radical polymerization [13,14], solid-state polymerization [15–17], chemical oxidative polymerization[18,19], emulsion polymerization [20], cationic polymerization [21],anionic polymerization [22,23], and oxidative polymerization via inor-ganic oxidants [24]. However, to date, these monomers have never

24 J.J. Senkevich / Thin Solid Films 556 (2014) 23–27

been chemical vapor deposited. Acenaphthylene is of particular interestsince it is photoluminescent (conductingwhen doped) and potentially ahigh-quality dielectric material respectively when polymerized. Poly-acenaphthylene (undoped) is also yellow in color, which could bevery useful for marketing purposes as a conformal coat. Of the afore-mentionedpolymerizationmethods, chemical oxidative polymerizationusing benzoyl peroxide is themost adaptable to a vacuum environmentwhere the technique has been called oxidative CVD (o-CVD) with theuse of inorganic oxidants [2]. That terminology will be used in thispaper here with the use of the organic oxidant benzoyl peroxide (III).

2. Experimental details

The CVD polymer thin films were deposited via a custom-builtvacuum system with separate vaporizers for the monomers and theoxidant. The monomers, acenaphthylene (A) (Tm 92 °C), acenaphthene(B) (Tm 93 °C),and indane (C) (b.p. 176 °C), were purchased fromAldrich and used as-received. They were vaporized under vacuum(base pressure 0.5–0.7 Pa) at 100 °C, 105 °C, and, 40 °C, respectively.Acenaphthylene had much tar contained within it, but the cleanchemical was easily vaporized with nearly all the tar left behind. Thedeposition time and temperature are shown in Table 1. The substratewas heated (or cooled) via a glycol loop with a substrate that hadserpentine-like channels made of brazed copper. The deposition cham-berwas cold-walled but the front-endwas heated via silicone heaters toat least the vaporization temperature of the monomers.

Benzoyl peroxide (III) (Aldrich) was dissolved in chloroform(Aldrich) and used as a 0.42-M solution. Chloroform was chosen as astable non-reactive solvent toward benzoyl peroxide and Viton, usedas both O-rings and tubing for direct liquid injection. Benzoyl peroxideis a solid and cannot be heated without an explosion in a closed system;therefore, it was dissolved in chloroform (direct liquid injection via aheated metering valve) then injected onto a hot manifold and flashevaporated. The vacuum system was pumped with a roughing pump/roots blower stack with an in-line foreline trap (organic vapor andparticulate)

The thickness and the optical properties of the polymer thin filmswere measured via a J.A. Woollam variable angle spectroscopicellipsometer (M-44 VASE) from 400 nm to 1000 nm, although the indi-ces of refraction (in plane and out of plane) were reported at 632.8 nm.Polyacenaphthylene (1A) is photoluminescent (conductive whendoped) and therefore strongly absorbing in the visible region. As aresult, optical properties could only be measured away from theirabsorption edge. In all cases, a biaxial Cauchy model was used withUrbach absorption.

Substrates used for deposition, and all characterizations werefloatzone (FZ) silicon b100 N double-side polished, n-doped with aresistivity of 2970–3210 Ω-cm (international wafer service). A Nexus870 FT-IR (Thermo Scientific) was used for infrared spectroscopymeasurements in transmission mode with a 4-cm−1 scan resolution,and 32 scans were undertaken both for background and samples scansafter purging at least 10 min with dry nitrogen.

3. Can themonomers be free radical polymerized? Considerations ofthe organic initiators

The intention in this study is touse anorganic oxidant, benzoylperoxide(III), to oxidatively polymerize acenaphthylene (A), acenaphthene (B),

Table 1Properties of the o-CVD polymers deposited in this study.

CVD polymer Deposition temperature Deposition time

polyacenaphthylene 125 °C 15 minpolyacenaphthene 125 °C 20 minpolyindane 35 °C 10 min

and indane (C) under low pressure vacuum conditions and at lowtemperatures ≤125 °C; however, two questions arise.

1) Can these monomers undergo a free radical polymerization?2) Does the oxidant itself, i.e., benzoyl peroxide, yield a polymer?

Certainly, acenaphthylene (A) has been polymerized in solution viaAIBN [13] and ABCN [14] both carried out in THF; however, these initi-ators are much more aggressive than vacuum-derived free radicalsmostly because their interaction time is greater in solution.

Two common free radical initiatorswere investigated t-butyl peroxide(I) and cumyl peroxide (II) using the method of i-CVD. Both of these freeradical initiators will readily polymerize acrylate monomers via i-CVDsuch as glycidyl-methacrylate at a filament temperature of N175 °Cand at a deposition temperature of ~35 °C [25]. This was confirmed ex-perimentally although the data is not presented here. I can be meteredthrough standard mass-flow controllers, whereas cumyl peroxide (II) isbest delivered via bubbling argon through it using t-butyl cumyl peroxide(Trigonox T, AkzoNobel), which is a liquid and more convenient to useunlike dicumyl peroxide, which is a solid.

It is very important tomatch the proper surface concentration of theinitiator with that of themonomer such that polymerizationswill occurand high molecular weight (MW) polymer is achieved.

• If toomuch initiator is present at the substrate surface, then polymer-ization will occur but low MW polymer will result and the initiatorwill be found in high concentration in the thin film.

• If not enough initiator is present, then the deposition rate will be veryslow and possibly not occur whatsoever.

The surface concentration is controlled by the flow of the chemical(initiator or monomer), the substrate temperature, and using the prop-er filament temperature in the case of i-CVD. These parameters need tobe optimized for each chemical used. In the case of I, it has a highvolatility (low MW) and therefore is not effective at polymerizing lowvolatility (high MW) monomers. As a result, I would be an effectiveinitiator for indane (C) at 35 °C because they have matched volatility,whereas cumyl peroxide (II) is better matched with acenaphthylene(A) and acenaphthene (B) at a substrate temperature of 125 °C. Whenthese experiments were undertaken in i-CVD mode at a filamenttemperature of 200 °C, no deposition occurred.

As a control sample, II was used to polymerize the low volatilityN-phenylmaleimide (NMP) monomer (173 g/mol, Tm 86 °C) at adeposition temperature of 125 °C and at a filament temperature of200 °C. NMP is an electron deficient monomer in contrast to A, B,and C and is therefore easy to polymerize via a free radical initiator.In summary, electron-rich monomers cannot be polymerized bythe i-CVD method, i.e., using a free radical initiator; this is summa-rized in Fig. 1.

A set of control samples were also undertaken, where I and II wereflowed over the filament array (i-CVD mode) heated to 200 °C and ata substrate temperature of 25 °C with no monomer was present.Again, no deposition was evident. The same sets of control experimentswere undertaken with benzoyl peroxide (III) by itself delivered viadirect liquid injection (DLI) in a chloroform solvent (0.42 M solution)both with heated filaments (~200 °C) and without heated filaments ata substrate temperature of 125 °C. In both cases, no deposition wasevident. If I, II, or III yielded a deposited thin film without the presenceof amonomer, then theywould not be appropriate initiators. The caveathere is that the chemistry cannot always be well predicted a priori,

Thickness nin-plane nout-of-plane k

118 nm 1.6758 1.6942 4.03E−03469 nm 1.6663 1.6595 3.03E−0374 nm 1.5799 1.5473 1.58E−03

Fig. 1. Free radical initiators I. t-butyl peroxide and II. Cumyl peroxide are not able to polymerize themonomers under studyA. acenaphthylene,B. acenaphthene, andC. indane, via the i-CVDmethod. They also do not yield polymers without the presence of the monomers, again via the i-CVD method.

25J.J. Senkevich / Thin Solid Films 556 (2014) 23–27

especially in the caseof oxidativeCVD. For example, t-butyl peroxybenzoateyields a polymer in i-CVD mode, which is acetone soluble.

4. Oxidative CVD of polyacenaphthylene, polyacenaphthene,polyindane, and polypyrole

Compared to depositing the parylene polymers, which are easilydelivered, self-initiated, and only are a one chemical component system,i-CVD and o-CVD are significantly more complicated to optimize prop-erly. In addition, in the current study, benzoyl peroxide (III) is flashevaporated via DLI using a chloroform solution. A peristaltic pumpwas purchased to control flow but just using a heated needle valvewith small diameter (1/8″ ID) Viton tubing was sufficient and morereliable for proof-of-concept experiments.

Benzoyl peroxide IIIhas aMWof 242 g/mol and has ameltingpointof 103–105 °C at which point it rapidly decomposes (explosively in aclosed system). At low temperatures b125 °C, used in this study, IIIwill not crack to form free radicals but may be used as an oxidant. IIIreadily oxidizes hydrocarbons [26,27] in the same way that ferricchloride works generating the radical cation intermediate, whichpropagates much like a free radical addition polymerization althoughthe mechanism is more complicated. It can be considered a powerfuloxidizing agent; for example, it will oxidize ferrous ions to ferric ions[28]. III is more suitable for oxidatively polymerizing acenaphthene(MW 154 g/mol) and acenaphthylene (MW 152 g/mol) rather thanindane (MW 118 g/mol) due to its low volatility. However, in eithercase, the amount of III added is critical as previously discussed, toachieve high MW polymer.

5. Polyacenaphthylene

The scheme for depositing the o-CVD films via benzoyl peroxide isshown in Fig. 2. A polyacenaphthylene (1A) film was deposited at

Fig. 2.Oxidative CVD using III. Benzoyl peroxide to polymerizeA,B, and C, yielding 1A, polyacenbond in 1A, and the position where the monomers have been polymerized has not been confir

125 °C for 15 min, yielding the infrared spectrum in Fig. 3 and its opticalproperties in Table 1.Acenaphthylene A retained its double bond as ev-idenced by the yellow color of the deposited film 1A on a glass slide andis therefore a conjugated polymer. If the polymerization mechanismwas via free radical, rather than oxidation, then the resulting polymer,polyacenapthylene, would not have been yellow since the doublebond would have opened up and conjugation lost. Fig. 3 exhibits somedifferences between itself and polyacenaphthylene (1A) as reported inthe literature; however, this is mostly to do with the synthetic methodsin the literature open up the double bond in acenaphthylene (A) to yieldpolyacenaphthene (1B). It is not clear even with the chemical oxidativepolymerization [18,19] used in the literature if the resulting polymer is1A. For example, Flowers and Miller [19] reported they made 1Abut stated a white powder was isolated more indicative of a non-conjugated polymer 1B. It is also evident that 1A is a highly symmetricalpolymer and not very ir active. In addition, the strongest chromophorein III is the carbonyl peak, and it is barely evident in Fig. 3. However,the peaks at 699 (Ar-H), 1032 (C-O), and 1218–1259 cm−1 (C-O) canbe associated with benzoyl peroxide [29]. The 118-nm thin film didnot dissolve in either acetone or isopropyl alcohol. It had an ave-rage index of refraction of 1.6819 at 632.8 nm and a reasonably lowextinction coefficient of 4.03e−3. It also exhibited some out-of-planeorientation of its naphthyl groups (positive birefringence). It alsoexhibited oxidation after heating it to 170 °C for 24 h with a loss ofthe peaks at 1032 cm−1 and 1492 cm−1 and the presence of newpeaks at 1270 cm−1 (C-O) and 1750 cm−1 (C = O). This oxidation ismuch like that of parylene N and parylene X [30], but possibly 1A ismore resilient since this temperature is rather aggressive.

6. Polyacenaphthene

Polyacenaphthene (1B), unlike polyacenaphthylene (1A), was trans-parent when deposited onto a glass slide, expected of a non-conjugated

aphthylene; 1B, polyacenaphthene; and 1C, polyindane. Note that there is no loss of doublemed.

3200 3000 2800 1500 1200 900 600

842

1032

802

1218

1259

699

1728

2960

3057

3027

1492

Wavenumber (cm-1)

Fig. 3. FT-IR spectrum of a polyacenaphthylene thin film deposited at 125 °C with athickness of 118 nm, shown in the as-deposited condition on double-side polished floatzone silicon.

26 J.J. Senkevich / Thin Solid Films 556 (2014) 23–27

polymer. Fig. 4 shows the infrared spectrum of 1B deposited at 125 °Cfor 20 min, yielding a 469-nm thin film. 1B in Fig. 4 exhibits a peak at1600 cm−1 unlike that of 1A. Also, the peaks at 2920–3057 cm−1,attributed to C-H stretches, are more prominent relative to the otherabsorption peaks in the spectrum much like compared B to A (themonomers) [29]. Fig. 4 also is much like the ir spectra presented in theliterature [12,23,24]. It should benoted the peak at 1725 cm−1 is attributedto benzoyl peroxide (III), with an excess possibly resulting in a lower MWpolymer being deposited.

The 1B film was aged for 1 month in laboratory conditions (20 °C,55% RH, overhead fluorescents), and the ir spectrum was run againresulting in a slight increase in the peak at 1725 cm−1, as shown inFig. 4. With an excessive amount of III present, previous films wereshown (not published) to be unstable resulting in much hydrolysis ofthe polymer films (−OH bonding at 3500 cm−1). 1B exhibited a slightlylower index of refraction of 1.6640 at 632.8 nm compared to 1A and aslightly lower extinction coefficient 3.03e−3 as shown in Table 1.Monomers A and B have nearly the same polarizability and MW, andtherefore their respective polymers should not have much differentindices of refraction unless their molecular packing is much different.The difference between the indices of refraction of 1A and 1B is likely

3200 2800 2400 2000 1600 1200 800

3029

1073

1262

1447 10

28

1600

1725

767

699

2920

3057

1492

Wavenumber (cm-1)

Fig. 4. FT-IR spectra of a polyacenaphthene thin film deposited at 125 °C with a thicknessof 469 nm. The top curve (black) in the as-deposited condition. The bottom curve (red),the same sample aged for 1 month under laboratory conditions.

within experimental error, i.e., not significant, since the experimentalmethod was not optimized.

It was very striking the contrast between the two polymer thin films,polyacenaphthylene (1A) and polyacenaphthene (1B) in terms of theircolor when they were deposited onto glass slides. More than likely, themonomers polymerize at the bridged positions. It has also been report-ed that 1Bwhen semi-crystalline has an orthorhombic unit cell [15,17].The current films did not exhibit strong optical anisotropy; however,the films were not post-deposition annealed to drive their crystalliza-tion [31]. Further, it is important to establish the synthesis or depositionmethod before extensive characterization should be undertaken withthese polymer thin film.

7. Polyindane

Indane (C) (also called indan or benzocyclopentane) is similar toacenaphthene (B), except it is liquid at room temperature and nottypically considered a monomer but instead a hydrocarbon solvent.The literature is very scare on the direct polymerization of C [32].Nametkin et al. reported in 1973 the polymerization of C with AlBr3yielding oligomers of an MW 280–300 g/mol [32]. In contrast, it ismuch more common to polymerize indene, which is indane with apolymerizable double bond [33–35].

The samemethodology used abovewith the polymerization ofA andB was used to deposit thin films of polyindane. Fig. 5 is the infraredspectrum of polyindane (1C) deposited at 35 °C for 10 min, yielding athin film 74 nm thick. The peak at 2962 cm−1 is attributed to C-Hbonding in the non-aromatic area of the repeat unit. There are six ofthese bonds compared to only three of the Ar-H bonds in the indanemonomer, and this is reflected in Fig. 5 in contrast to 1A and 1B. Thepeak at 805 cm−1 is associated with ortho-disubstituted benzene andshifted from the monomer at 751 cm−1. The peaks at 1024, 1086, and1262 cm−1 are also present in the monomer; however, the peaksaround 1460 cm−1 in the monomer spectrum are absent in the poly-mer. There are also no peaks present associated with the carbonyl(C = O) stretch around 1725 cm−1.

As shown in Table 1, 1C exhibited a lower index of refraction com-pared to 1A and 1B with an average index of refraction of 1.5690 at632.8 nm. Its lower extinction coefficient is a result that it is thinner.Since literature on polyindane is very scarce, it is not clear where onthe indane monomer the polymerization is being undertaken. Quitepossibly, amechanistic analysis of the polymerization, i.e., the formationof the cation radical intermediate, could elucidate this mystery. Also, if athicker polyindane film was deposited, would that result in different

3200 3000 2800 1400 1200 1000 800 600

604

805

1024

1086

2962

3030

3062

1262

Wavenumber (cm-1)

Fig. 5. FT-IR spectrum of a polyindane thin film, shown in the as-deposited condition,deposited at 35 °C with a thickness of 74 nm.

27J.J. Senkevich / Thin Solid Films 556 (2014) 23–27

optical properties? Films less than 100 nm are generally consideredultra-thin films and often exhibit different optical properties (index ofrefraction and anisotropy) compared to films 100–1000 nm thick [36].

It should be noted that polypyrole was deposited by the methodpresented here but much benzoyl peroxide was incorporated into thethin film because of the largeMWdifference between benzoyl peroxide242 g/mol compared to that of pyrrole 67 g/mol. Interestingly, thebenzoyl peroxide that was deposited away from the heated areas ofthe deposition chamber turned purple with the reaction of the pyrrole.Polypyrrole has been made in the literature with benzoyl peroxide insolution via emulsion polymerization with a similar color [37]. Also, itshould be noted that the polymerization of tetralin and p-xylene wasattempted via the o-CVD method using benzoyl peroxide and aftermultiple attempts no thin film could be deposited with either chemical.Quite possibly, the cation free radical intermediate is not stable in thesemolecules.

8. Conclusions

A method is presented to polymerize inexpensive and readily avail-able electron-rich monomers via oxidative CVD, which overcomes thedrawbacks of using ferric chloride and molecular bromine as oxidants,namely, that they are highly corrosive toward humans and vacuumsystems. Benzoyl peroxide is however explosive and needs to be deliv-ered via direct liquid injectionwith aflash evaporator. That said, one canenvision the use of other volatile, non-corrosive, non-explosive organicoxidants to polymerize these and other electron-richmonomers, whichwould enable the commercialization of these attractive polymers thatare both high-quality dielectrics and conjugated polymer systems,which once doped, would be conductive.

The method presented here also overcomes the issue of inexpensivemonomer availability. It is very difficult to present the need to customsynthesize monomers because of regulation and cost issues. Themonomers here are either coal-tar derivatives (acenaphthylene andacenaphthene) or petroleum stock chemicals (indane), which makestheir cost and availability almost trivial.

References

[1] J.J. Senkevich, P.-I. Wang, Chem. Vap. Depos. 15 (2007) 91.[2] W.E. Tenhaeff, K.K. Gleason, Adv. Funct. Mater. 18 (2008) 979.[3] B.P. Carrow, H. Bakhru, P.-I. Wang, Y. Chen, J.J. Senkevich, Chem. Vap. Depos. 12

(2006) 239.[4] F. Intranuovo, E. Sardella, P. Rossini, R. d'Agostino, P. Favia, Chem. Vap. Depos. 15

(2009) 95.[5] H.G. Pryce Lewis, D.J. Edell, K.K. Gleason, Chem. Mater. 12 (2000) 3488.[6] H. Nandivada, H.-Y. Chen, L. Bondarenko, J. Lahann, Angew. Chem. Int. Ed. 45 (2006)

3360.[7] W.F. Gorham, J. Polym. Sci. A1 4 (1966) 3027.[8] J.J. Senkevich, Chem. Vap. Depos. 17 (2011) 235.[9] K.K.S. Lau, H.G. Pryce Lewis, S.J. Limb, M.C. Kwan, K.K. Gleason, Thin Solid Films 395

(2001) 288.[10] D. Bhattacharyya, K.K. Gleason, Chem. Mater. 23 (2011) 2600.[11] W.R. Dolbier, W.F. Beach, J. Fluor. Chem. 122 (2003) 97.[12] H. Liu, J. Xu, S. Pu, M. Luo, Z. Le, Mater. Lett. 61 (2007) 1392.[13] J. Springer, J. Schmelzer, T. Zeplichal, Chromatographia 13 (1980) 164.[14] I.C. McNeill, M.H. Mohammed, Polym. Degrad. Stab. 55 (1997) 191.[15] V.C. Bastron, H.G. Drickamer, J. Solid State Chem. 3 (1971) 550.[16] C.S. Hsia Chen, J. Polym. Sci. 62 (1962) S38.[17] A. Usanmaz, N. Baytar, J.M.S. Pure Appl. Chem. A35 (1998) 161.[18] H. Agbaje, J. Springer, Eur. Polym. J. 22 (1986) 943.[19] R.G. Flowers, H.F. Miller, J. Am. Chem. Soc. 69 (1947) 1388.[20] C.-Y. Chen, I. Piirma, J. Polym. Sci. Polym. Symp. 65 (1978) 55.[21] P. Giusti, G. Puce, F. Andruzzi, Makromol. Chem. 98 (1966) 170.[22] J. Moacanin, A. Rembaum, Polym. Lett. 2 (1964) 979.[23] J. Springer, T. Win, Makromol. Chem. 179 (1978) 29.[24] H.-N. Cho, S.-K. Choi, J. Polym. Sci. A Polym. Chem. 25 (1987) 15.[25] M. Gupta, K.K. Gleason, Thin Solid Films 515 (2006) 1579.[26] E.V. Titov, Y.N. Alaev, L.V. Luk'yanenko, Theor. Exp. Chem. 21 (1985) 596.[27] G.S. Hammond, U.S. Nandi, J. Am. Chem. Soc. 83 (1961) 1213.[28] F.T. Wall, T.J. Swoboda, J. Am. Chem. Soc. 71 (1949) 919.[29] NIST Standard Reference Database 69: NIST Chemistry WebBook2011.[30] J.J. Senkevich, Chem. Vap. Depos. 17 (2011) 204.[31] J.J. Senkevich, S.B. Desu, V. Simkovic, Polymer 41 (2000) 2379.[32] N.S. Nametkin, E.Sh. Finel'shtein, E.B. Portnykh, V.M. Vdovin, V.D. Oppengeim, Dokl.

Akad. Nauk SSSR 210 (1973) 858.[33] F.J.B. Brum, F.N. Laux, M.M.C. Forte, Des. Monomers Polym. 16 (2013) 291.[34] M. Marref, R. Meghabar, M. Belbachir, Ann. Chim. 35 (2010) 73.[35] H.F. Naguib, S.M. Mokhtar, N.Z. Khalil, M.Z. Elsabee, J. Polym. Res. 16 (2009)

693.[36] J.J. Senkevich, J. Vac. Sci. Technol. A 18 (2000) 2586.[37] C. Saravanan, R.C. Shekhar, S. Palaniappan, Macromol. Chem. Phys. 207 (2006)

342.