Thin Solid Films 603 (2016) 8–13

Contents lists available at ScienceDirect

Thin Solid Films

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

Properties of silicon nitride thin overlays deposited on opticalfibers — Effect of fiber suspension in radio frequency plasma-enhancedchemical vapor deposition reactor

M. Śmietana a,⁎, M. Dominik a, M. Myśliwiec a, N. Kwietniewski a, P. Mikulic b, B.S. Witkowski c, W.J. Bock b

a Institute of Microelectronics and Optoelectronics, Warsaw University of Technology, Koszykowa 75, Warsaw 00-662, Polandb Centre de Recherche en Photonique, Université du Québec en Outaouais, 101 rue Saint-Jean-Bosco, Gatineau, J8X 3X7, Québec, Canadac Institute of Physics, Polish Academy of Sciences, Al. Lotników 32/46, Warsaw 02-666, Poland

⁎ Corresponding author.E-mail address: M.Smietana@elka.pw.edu.pl (M. Śmie

http://dx.doi.org/10.1016/j.tsf.2016.01.0460040-6090/© 2016 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 4 March 2015Received in revised form 13 November 2015Accepted 22 January 2016Available online 25 January 2016

This work discusses the effect of sample suspension in radio frequency plasma-enhanced chemical vapor depo-sition process on properties of the obtained overlays. Silicon nitride (SiNx) overlayswere deposited on flat siliconwafers and cylindrical fused silica opticalfibers. The influence of the suspension height andfiber diameter on SiNx

deposition rate is investigated. It has been found that thickness of the SiNx overlay significantly increases withsuspension height, and the deposition rate depends on fiber dimensions. Moreover, the SiNx overlays werealso deposited on long-period gratings (LPGs) induced in optical fiber. Measurements of the LPG spectral re-sponse combined with its numerical simulations allowed for a discussion on properties of the deposited overlay.The measurements have proven higher overlay deposition rate on the suspended fiber than on flat Si waferplaced on the electrode. Results of this work are essential for precise tuning of the functional properties of newgenerations of optical devices such as optical sensors, filters and resonators, which typically are based on opticalfibers and require the overlays with well defined properties.

© 2016 Elsevier B.V. All rights reserved.

Keywords:Thin filmSilicon nitridePlasma-enhanced chemical vapor depositionLong-period gratingsOptical fiber

1. Introduction

Thin films deposited on various materials and shapes are requestedby number of novel applications, including fabrication on optical,electrical and chemical devices. There are a lot of applications, wherethe thickness or optical properties are critical for device operation.In those cases properties of the films must be controlled down tonanometric scale, and what's even more challenging — on objectswith complex shapes. Thin films can be depositedwith variousmethodswhich enable uniform deposition independently on substrate shape,such as the following: a self assembling monolayer deposition methodwhich relies on liquid precursors [1], atomic layer deposition, which inturn employs gas precursors [2] or layer-by-layer deposition method[3]. Unfortunately, all these methods are relatively time-consumingand limited in terms of materials that can be deposited with them.

In this experiment we investigate a capability of radio-frequencyplasma-enhanced chemical vapor deposition (RF PECVD) in parallelelectrode configuration for uniform deposition of thin overlays on di-electric substrates of cylindrical shape which are suspended in plasma.The method is well established for deposition on flat wafers. The RFPECVD offers set of advantages, including, e.g., low-cost film fabrication,high efficiency of deposition and capability for deposition of graded-

tana).

refractive-index films or as a stack of nano-films each with differentproperties [4]. The growth of the film is due to activation of the gas-phase precursors in a glow-discharge (plasma) environment, wherethe chemical reactions are activated by plasma and take place over thesubstrate as well as at the substrate. In PECVD method, which belongsto a group of chemical vapor deposition methods (CVD), the amountof deposited film on a sample depends mainly on effectiveness ofplasma-enhanced decomposition of gases and not only onmass locatedin the reaction chamber, but also on sample shape, sample location inchamber (plasma), particle (ion) flux, material of the sample and pa-rameters of the process, such as gas composition, pressure, power, andtime. The uniformity of silicon carbide coating deposited with the RFPECVD method on complex shapes placed on the cathode has alreadybeen discussed [5]. In our previous work on silicon nitride (SiNx) filmsdeposited on elevated silicon substrates, we have shown that the sam-ple orientation in the PECVD chamber has strong influence on thicknessand on optical properties of the films [6]. These amorphous [7] films aredense, free from defects and pores, show good adhesion to Si and SiO2

substrates, good chemical stability, and excellentmechanical properties[8,9].What ismore, the refractive index (n) of SiNxfilms can be adjustedin wide range by tuning deposition conditions [6]. These properties arevery important in fabrication of optical waveguides and planar opticalsystems [8–12] and for designing of novel photonic and optoelectronicdevices [8,13,14]. Application of SiNx as a coating for LPGs will allowfor manufacturing fiber sensing structures in low temperature process.

9M. Śmietana et al. / Thin Solid Films 603 (2016) 8–13

These dense anduniformfilmson sensor structure limit risk of deforma-tion or degradation of an optical fiber. The SiNx films simultaneouslywith fiber protection, can play a role of coatingswhich increase sensitiv-ity of a LPG sensor to refractive index, bending, pressure or temperature.

In our previous work on silicon nitride (SiNx) films deposited onelevated silicon substrates, we have shown that the sample orientationin the PECVD chamber has strong influence on thickness and on opticalproperties of the films [6]. We have previously discussed an applicationof the SiNx films as overlays for several types of optical fiber sensors,where nanocoatingmodifies the sensor response on external conditions[15–17]. For such optical sensing application an exact determination ofoverlay properties on fused silica optical fibers is crucial. In thisworkwefocused on investigations of SiNx thin films deposited on the optical fi-bers, with no sample rotation applied during the process. These investi-gations are essential, e.g., for precise tuning of the functional propertiesof new generations of optical devices such as optical sensors, filters andresonators, which typically are based on optical fibers and require theproperties of the overlays to be well defined. For the needs of thiswork the SiNx overlays were also deposited on the long-period gratings(LPGs) induced in optical fiber. The LPG is a periodic modulation ofrefractive index within a core of the fiber resulting in appearance of aset of resonances in its spectral response [18]. The measurements ofthe response of the gratings combined with the numerical simulationswill allow to conduct an analysis of overlay properties deposited onthe LPG [19–21].

2. Experimental details

SiNx thin films were deposited on p-type b100N silicon substrates(ρ = 1–10 Ω cm) as well as on optical fibers. Fused silica optical fiberswith different diameters (Corning® SMF-28™ single-mode opticalfiber and polymer-clad silica multimode fiber, with core diameter of125 μm and 400 μm, respectively) after mechanical and chemicalremoval of polymer overlays were placed in Oxford Plasmalab 80 Plussetup on stainless steel holder and their height (h) was changed from2.5 to 8 mm (Fig. 1). The fiber length was 10 cm for both fiber samples.The reactor used in this experiment is a parallel-plate reactor withgraphite electrode (240 cm in diameter), where wafers are placed onthe heated substrate and the source gases are introduced from the topof the reactor and pumped away around the bottom electrode. The op-erating frequencywas 13.56MHz. Other deposition process parameterswere similar to those applied in [5], i.e., SiH4 and NH3 flow of 150 and50 sccm, respectively, pressure 120 Pa, electrode temperature 200 °C,RF generator power 50 W, and deposition time was up to 15 min.

Fig. 1.A schematic representation of a holder used for suspension of thefiber samples in the reacwafer, (3) stainless steel holder, and (4) flattening 3-inch silicon wafer. The holder, as seen fro

Simultaneously, the films were also deposited on reference siliconwafer placed on the holder's surface. Properties of the SiNx films depos-ited on the Si substrates, i.e., thickness, n and extinction coefficient (k),were investigated with Horiba Jobin-Yvon UVISEL spectroscopicellipsometer, according to procedure described in details in [5].Thickness of the overlay on optical fibers was investigated after fiber'scleaving and observations of their cross-section using Hitachi S-3400Nscanning electron microscope (SEM) and operating at 13 kV and 15 kV.

The LPGs used in this experiment were fabricated in the sameCorning® SMF-28™ optical fiber as the one used for described aboveoverlay thickness investigation. The LPGs were manufactured using anelectric-arc method with a period of 500 μm [22]. Transmission spectraof the LPGs in wavelength range λ = 1160–1660 nm were comparedbefore and after the SiNx deposition, as well as for the LPG immersionin liquids of different n (nD = 1.33 to 1.41). Optical transmission mea-surements were performed using Agilent 83437A broadband lightsource and an Agilent 86142B optical spectrum analyzer. Numericalsimulations of the overlay deposition on the LPG have been performedusing Optigrating v. 4.2 software by Optiwave Systems [22].

3. Results and discussion

3.1. Evolution of the overlay thickness with the optical fiber suspension

The most reasonable way to estimate thickness of the SiNx nano-overlay on the fiber surface seems to be microscopic investigations oftheir cross-section. The SEM images with estimated thickness of thecoating for different sample diameter and for two selected suspensionheights are shown in Fig. 2. It can be clearly seen that thickness of theoverlays significantly increases with the distance between the fiberand the holder surface. The increase is higher for the thinner fiber(Ø = 125 μm), and when it's placed 8 mm above the holder's surface,the thickness of the overlay is over fourfold thicker than the for thefilm deposited on the siliconwafer placed directly on the holder surface.For samples suspended in proximity to the holder surface (below4 mm), a visible non-uniformity of the overlay around the fiber hasbeen found.

In Fig. 3, the averaged thickness of the overlays deposited on boththinner (Ø = 125 μm) and thicker (Ø = 400 μm) fused silica fibers forprecisely adjusted suspension height in plasma reactor is compared. Asignificant difference between kinetics of the overlay deposition forthe two fibers has to be noted. The thinner fiber promotes higher depo-sition rates than those measured for the thicker fiber when suspendedin plasma. Moreover, the increase in deposition rate is higher for both

torwhere a) shows the holder from the top and (1) represents a fiber, (2) reference siliconm the side with marked suspension adjustable height (h), is shown in b).

Fig. 2. SEM images of the SiNx overlays deposited on optical fiber sample suspended at different height, where a) and b) show images for optical fiber with 125 μm in diameter— 3 and8 mm suspension, respectively; c) and d) for 400 μm in diameter fiber — 3 and 8 mm suspension, respectively. The deposition was 15-min long.

10 M. Śmietana et al. / Thin Solid Films 603 (2016) 8–13

these fibers than for the deposition rate previously observed on Si wafersuspended on stainless steel bar [6].

The PECVDmethod is based on activation gases in plasma. Activatedparticles diffuse towards the sample surface and react there whichresults in deposition of a film. The RF PECVD is typically a low-temperature process, where the decomposition of gases (in our caseSiH4 and NH3) and reactions between the decomposition products areenhanced by plasma [23]. Simplified reaction scheme resulting in

Fig. 3. Influence of the distance between fiber and holder surface on average thickness ofthe deposited film for two fused silica optical fiberswith different diameters suspended inplasma. The increase factor is calculated in reference to the film thickness deposited on Siwafer placed on the holder surface (246 nm).

formation of SiNx on the substrate can be expressed in three mainsteps [24]:

1. Gas phase dissociation of NH3 ((1)) and SiH4 ((2));

NH3 þ e→NHb þ H3−b þ e ð1Þ

SiH4 þ e→SiHa þH4−a þ e ð2Þ

2. formation of a-Si:H-like film on the substrate by SiHa and NHb

radicals; and3. film densification and formation of Si\\N bonds by the following:

- cross-linking at elevated substrate temperature to break Si-H andN-H ((3)),

Si‐Hþ N‐H→Si‐NþH2 ð3Þ

- the excessive incorporation of NHb radicals, which breaks and re-places the existing Si\\H bonds (at elevated substrate temperature)((4)),

Si‐Hþ NHb→Si‐Nþ Hbþ1 ð4Þ

- the excessive incorporation of SiHa radicals, which in turn breaksand replaces the existing N\\H bonds ((5)),

N‐Hþ SiHa→Si‐Nþ Haþ1 ð5Þ

It can be seen that there are two types of reactions: homogeneous(taking place in plasma) and heterogeneous at the substrate surface,where SiHa and NHb radicals interact [25,26]. The later one is dominates

Fig. 4. Influence of SiNx overlay deposition and of increase in external n (nD = 1.333 to1.421) on transmission spectrum of the LPG. Arrows show transition of LP05 and LP06modes with external n.

11M. Śmietana et al. / Thin Solid Films 603 (2016) 8–13

in the investigated case. Moreover, it is known that RF generator powerand pressure in the reaction chamber are the main process parametersdetermining (a, b) the product ratio [26].

The main factors affecting in this work the film deposition rate aredistance between the sample and the electrode (h), the shape of sample,and temperature at the sample. Both experiments of deposition on fi-berswith different dimensionswere performed in the same place insidethe reactor and using the same holder. That is why we can assume thatthere is no influence of the sample position in the reactor or change inshape of the holder on the result of the experiment.

3.1.1. Distance between the sample and the electrodeWhen the samples are placed at different h, there is a different

distance between the electrodes and the sample. During the RF PECVDprocess the mentioned radicals and molecules are accelerated byelectric field induced between the electrodes. As shown in this paper,the increase in deposition rate (thickness of the deposited film) withan increase in h is the most probably caused by the increased numberof radicals reaching the samples during one frequency cycle of the RFgenerator. When the sample is placed higher, more silicon and nitrogencontaining radicals reach the sample surface than when the sample isplaced at lower h. Particles need a certain time to make the distancebetween the area of plasma and the sample. This time depends mainlyon this distance, as well as on power and frequency of the generator,and to a lower degree on gas flow, pressure and process temperature.As the sample is placed higher and closer to plasma area the particleshave less distance to travel, and thus the deposition rate increases.High deposition rate for the sample located higher in the reactor mayalso be caused by a higher density of particles at the top of the reactor,where gas inlet is located. The higher density of electrons results in anincrease of the probability of dissociation and ionization of gas, whichin turn leads to an increase of deposition rate [27]. It has been shownthat more ions are introduced to reaction chamber the higher is deposi-tion rate of SiNx [24]. Similar effects have also beenmodeled when flowof reactive gases [28] or shape of gas inlet [29] were changed.

3.1.2. Shape of the sampleAnother factor influencing the deposition rate of the layers in this

experiment was the shape of the samples. We have shown that onthinner fiber the SiNx overlay is deposited with higher rate (Fig. 3).This phenomena is likely to be caused by changes in the ion flux. Theshape of the sample affects the surrounding plasma and disturbs pathsof ions towards the cathode. The sharp edges of the thinner silica fiberand heterogeneity of species distribution leads to a higher ion flux,which in turn increases deposition rate [30].

3.1.3. Temperature at the sampleThe last factor affecting thedeposition rate of SiNx in this experiment

is temperature at the sample. The active gases used in the PECVD pro-cesses easily react with each other or dissociate. When the reactiongases are at a high temperature, they dissociate thermally and it leadsto a compositional change of radicals in the gas phase, e.g., change of ynumber in SiH4 − y(NH2)y radical. In order to prevent this effect, thehighest temperature in the reaction zone is induced at the substratelevel, i.e., at the bottom heated electrode [23]. That is why the furtherfrom the heated electrode the sample is located in the reactor, thelower is the temperature. In this experiment the fibers are placed higherthan the reference Si wafer that lies directly on the holder. It has beenshown that the deposition rate of the SiNx films strongly depends onprocess temperature [6,31,32]. Kim et al. [31] reported that in the tem-perature range between 200 and 400 °C, the deposition rate of the SiNx

decreaseswith temperature, whereas in the range of 700–800 °C the in-crease in temperature is followed by the increase in deposition rate [32].We have performed the experiment at lower temperature (200 °C),therefore the increase in distance between the heated electrode andthe sample, followed by decrease in temperature, should rather induce

a decrease in thedeposition rate [6,31]. However, it has been also shownbyMehdipour et al. [33] that theflux of particles in PECVD process heatsup the sample surface. Therefore, when the sample is placed higherthe temperature at its surface can be higher than that at the sample-substrate holding platform. The increase in temperature results in alarger kinetic energy of particles which in turn leads to an increase ineffectiveness of the ion-induced dissociation. Therefore, as a result ofhigher ion energies and much larger particles fluxes to the samplewhen it is suspended in the reactor, the deposition rate increases. Thehigher sample is suspended, the observed phenomena aremore effective.

3.2. Investigation of the LPGs with SiNx nano-overlays

In order to prove a higher deposition rate observed on fiberssuspended in the RF PECVD reactor than on the substrate placed directlyon the electrode, we have deposited the SiNx overlays also on the LPGs.It has been shown that thanks to coupling between the core mode(LP01) and a series of the cladding modes (LP0m, where m = 2,3,4,…)a spectral response of the LPG highly depends on external n [35]. Theresponse also changes when an overlay with a defined thickness andoptical properties, i.e. n and k, is deposited on the LPG [21]. An increasein external n or in deposition of a coating induces an increase in effectiverefractive index (neff) of the cladding modes and this effect is seen as ashift of the resonances in transmission spectrum towards a lowerwave-length [34]. In case of thin overlay deposition, the wavelength shifthighly depends on properties of the overlay. When it is assumed thatoptical properties in a certain thickness range are independent on over-lay thickness [35], the change in overlay thickness can be monitored bytracing a shift of the resonances [36].

In this experiment the LPG has been suspended at h = 8.35 mm.Keeping in mind that at this height the deposition rate reaches4.14 × 16.4 ≈ 68 nm/min, in order to deposit an overlay thickness of100 nm, the deposition time was set to 88 s. Both thickness and n forsuch a film, as determined on the reference Si wafers are 28.7 nm andover 1.8, respectively [37]. It can be seen in Fig. 4, that the resonancesvisible for the LPG with no overlay, after the SiNx deposition experi-enced a spectral shift and the one at λ ≈ 1550 nm has split into two.The spectral shift is induced by an increase in neff of the claddingmodes when a high-n overlay is deposited [36]. The separation of theresonances in turn is an effect of asymmetrical overlay depositionaround the fiber [38]. After immersion of the LPG in water, theresonance has shifted towards shorter wavelength by over 50 nm.Comparing this result to a similar LPG with no overlay [22], the shift isover fivefold higher. It is also shown in Fig. 4, that when external n in-creases, the resonance experiences further shift and another resonanceappears in a higher wavelength range. The effect, when resonances

Fig. 5. Simulation of influence of overlay thickness on (a) neff ofmodes LP01 to LP08 at λ=1550 nmand (b) resonancewavelength in the range of 1500–1600 nmwhen LPG is surroundedby air (n = 1) and water (n = 1.3180@λ = 1550 nm). Transition of LP02, LP03 and LP04 to the overlay with an increase of external n is marked in b).

12 M. Śmietana et al. / Thin Solid Films 603 (2016) 8–13

replace the original ones at lower wavelength range is known as modetransition [39]. At such conditions one of the cladding modes of thelowest order, i.e., LP02, starts to propagate in the overlay and othermodes reorganize while heading towards an original mode structure,i.e., as shown in Fig. 4 — LP06 to LP05, LP05 to LP04. The transition condi-tions are mainly dependent on overlay properties, i.e., its thickness andn, as well as on external n. The simulations of this effect for modes LP01to LP08, assuming n= 1.9 [6] for higher thickness of the overlay, and foroverlay thickness up to 1400 nm are shown in Fig. 5a. Higher the exter-nal n is, the transition takes place for lower thickness of the overlay.Since in this experiment the resonance shift induced by immersion inwater is significant and since for a slightly higher external n a new res-onance appears in higher wavelength range, the overlay must be thickenough to support the transition. According to the simulations shownin Fig. 5b, the increase in a shift between water and air takes place forSiNx thickness close to 120, 700 and 1300 nm, where respectivelyLP02, LP03 and LP04 start to propagate in the overlay. The results provethat the thickness of the overlay deposited on the fiber must be over100 nm in order to fulfill the transition conditions.

4. Conclusions

Properties of the SiNx films deposited with the RF PECVD methodhighly depend on placement of the samples in the plasma reactor, aswell as on dimensions of the sample. We have shown that the thicknessof thefilms deposited on the opticalfibers suspended above the electrodeincreaseswith suspension height and can be over fourfold higher than forthe films deposited on the samples placed close to the electrode. The in-crease in the deposition ratewith suspension height can be caused by theincreased number of radicals reaching the samples during one frequencycycle of the RF generator, by higher density of particles at the top of thereactor as well as by increased substrate temperature due to the flux ofthe particles when it is suspended. Moreover, sharp edges of the thinnersilica fiber and heterogeneity of the species distribution lead to a higherion flux at its surface, which in turn increase the deposition rate. The ef-fect of strong variation of the properties with shape and placement inthe RF PECVD deposition chamber must be taken into considerationwhen application of the method and thin films is foreseen for coatingoptical-fiber-based devices and other devices with complex shapes. Webelieve that results shown in this paper can be a very good introductionto studies on similar layers such as SiO2, SiON deposited with PECVDusing the same gases such as SiH4 and NH3.

Acknowledgments

The authors gratefully acknowledge support for this work from theNational Centre for Research and Development of Poland within

LIDER/03/16/L-2/10 project, and the Polish National Science Centrewithin 2011/03/D/NZ7/02054 project.

References

[1] K. Ariga, J.P. Hill,M.V. Lee, A. Vinu, R. Charvet, S. Acharya, Challenges and breakthroughsin recent research on self-assembly, Sci. Technol. Adv. Mater. 9 (2008) 014109.

[2] E. Graugnard, D.P. Gaillot, S.N. Dunham, C.W. Neff, T. Yamashita, C.J. Summers,Photonic band tuning in two-dimensional photonic crystal slab waveguides byatomic layer deposition, Appl. Phys. Lett. 89 (2006) 181108.

[3] S. Korposh, S.W. James, S.-W. Lee, S. Topliss, S.C. Cheung, W.J. Batty, R.P. Tatam, Fiberoptic long period grating sensors with a nanoassembled mesoporous film of SiO2

nanoparticles, Opt. Express 18 (2010) 13227–13238.[4] M. Śmietana, M. Koba, R. Różycki–Bakon, Stack of PECVD silicon nitride nano-films

on optical fiber end-face for refractive index sensing, Proc. SPIE 9157 (2014) 91575F.[5] H. Anma, J. Toki, T. Ikeda, Y. Hatanaka, Uniform deposition of SiC thin films on

plastics surfaces, Vacuum 59 (2000) 665–671.[6] M. Śmietana, R. Mroczyński, N. Kwietniewski, Effect of sample elevation in RF

PECVD reactor on optical properties and deposition rate of silicon nitride thinfilms, Materials 7 (2014) 1249–1260.

[7] K.-C. Lin, S.-C. Lee, The structural and optical properties of a-SiN:H prepared byplasma enhanced chemical-vapor deposition, J. Appl. Phys. 72 (1992) 5474–5482.

[8] W.H.P. Pernice, M. Li, D.F.G. Gallagher, H.X. Tang, Silicon nitride membrane photon-ics, J. Opt. A Pure Appl. Opt. 11 (2009) 114017.

[9] G. Subhash, P. Hittepole, S. Maiti, Mechanical properties of PECVD thin ceramicfilms, J. Eur. Ceram. Soc. 30 (2010) 689–697.

[10] L. Martinu, D. Poitras, Plasma deposition of optical films: a review, J. Vac. Sci.Technol. A 18 (2000) 2619–2645.

[11] N. Daldosso, M. Melchiorri, F. Riboli, M. Girardini, G. Pucker, M. Crivellari, P. Bellutti,A. Lui, L. Pavesi, Comparison among various Si3N4 waveguide geometries grownwithin a CMOS fabrication pilot line, J. Lightwave Technol. 22 (2004) 1734.

[12] H. Haeiwa, T. Naganawa, Y. Kokubun, Wide range center wavelength trimming ofvertically coupled microring resonator filter by direct UV irradiation to SiN ringCore, IEEE Photon. Technol. Lett. 16 (2004) 135–137.

[13] A. Sinibaldi, E. Descrovi, F. Giorgis, L. Dominici, M. Ballarini, P. Mandracci, N. Danz, F.Michelotti, Hydrogenated amorphous silicon nitride photonic crystals for improved-performance surface electromagnetic wave biosensors, Biomed. Opt. Express 10(2012) 2405–2410.

[14] D. Li, F. Wang, D. Yang, D. Que, Electrically tunable electroluminescence from SiNx-based light-emitting devices, Opt. Express 20 (2012) 17359–17366.

[15] M. Śmietana, W.J. Bock, P. Mikulic, J. Chen, Pressure sensing in high-refractive-indexliquids using long-period gratings nanocoated with silicon nitride, Sensors 10(2010) 11301–11310.

[16] M. Śmietana, W.J. Bock, P. Mikulic, Temperature sensitivity of silicon nitridenanocoated long-period gratings working in various surrounding media, Meas.Sci. Technol. 22 (2011) 115203.

[17] M. Śmietana, D. Brabant, W.J. Bock, P. Mikulic, T. Eftimov, Refractive-index sensingwith inline core-cladding intermodal interferometer based on silicon nitride nano-coated photonic crystal fiber, J. Lightwave Technol. 30 (2012) 1185–1189.

[18] A.M. Vengsarkar, P.J. Lemaire, J.B. Judkins, V. Bhatia, T. Erdogan, J.E. Sipe, Long-periodfiber gratings as band-rejection filters, J. Lightwave Technol. 14 (58–65) (1996).

[19] M. Śmietana, M. Koba, P. Mikulic, W.J. Bock, Measurements of reactive ion etchingprocess effect using long-period fiber gratings, Opt. Express 22 (2014) 5986–5994.

[20] M. Śmietana, M. Koba, P. Mikulic, W.J. Bock, Tuning properties of long-periodgratings by plasma post-processing of their diamond-like carbon nano-overlays,Meas. Sci. Technol. 25 (2014) 114001.

[21] I. Del Villar, I.R. Matias, F.J. Arregui, P. Lalanne, Optimization of sensitivity in long pe-riod fiber gratings with overlay deposition, Opt. Express 13 (2005) 56–69.

[22] M. Śmietana, M. Myśliwiec, P. Mikulic, B.S. Witkowski, W.J. Bock, Capability for finetuning of refractive index sensing properties of long-period gratings by atomic layerdeposited Al2O3 overlays, Sensors 13 (2013) 16372–16383.

13M. Śmietana et al. / Thin Solid Films 603 (2016) 8–13

[23] M. Konuma, Film Deposition by Plasma Techniques, Springer-Verlag, Berlin, 1992.[24] Y. Wan, K.R. McIntosh, A.F. Thomson, Characterisation and optimisation of PECVD

SiNx as an antireflection coating and passivation layer for silicon solar cells, AIPAdv. 3 (2013) 032113.

[25] X. Xu, Q. He, T. Fan, Y. Jiang, L. Huang, T. Ao, C. Ma, Hard and relaxed a-SiNxHy filmsprepared by PECVD: structure analysis and formation mechanism, Appl. Surf. Sci.264 (2013) 823–831.

[26] D.L. Smith, A.S. Alimonda, C.-C. Chen, S.E. Ready, B. Wacker, Mechanism of SiNxHydeposition from NH3-SiH4 plasma, J. Electrochem. Soc. 137 (1990) 614–623.

[27] H. Huang, K.J. Winchester, A. Suvorova, B.R. Lawn, Y. Liu, Y.Z. Hu, J.M. Dell, L. Faraone,Effect of deposition conditions onmechanical properties of low-temperature PECVDsilicon nitride films, Mater. Sci. Eng. A 435–436 (2006) 453–459.

[28] L. Sansonnens, J. Bondkowski, S. Mousel, J.P.M. Schmitt, V. Cassagne, Development ofa numerical simulation tool to study uniformity of large area PECVD film processing,Thin Solid Films 427 (2003) 21–26.

[29] Y.-J. Kim, J.-H. Boo, B. Hong, Y.J. Kim, Effects of showerhead shapes on the flow fieldsin a RF-PECVD reactor, Surf. Coat. Technol. 193 (2005) 88–93.

[30] N. Nelson, R.T. Rakowski, J. Franks, P. Woolliams, P. Weaver, B.J. Jones, The effect ofsubstrate geometry and surface orientation on the film structure of DLC depositedusing PECVD, Surf. Coat. Technol. 254 (2014) 73–78.

[31] B. Kim, J.Y. Park, K.K. Lee, J.G. Han, Temperature effect on deposition rate of siliconnitride films, Appl. Surf. Sci. 252 (2006) 4138–4145.

[32] P. Temple-Boyer, C. Rossi, E. Saint-Etienne, E. Scheid, Residual stress in low pressurechemical vapor deposition SiNx films deposited from silane and ammonia, J. Vac. Sci.Technol. A 164 (1998) 2003–2007.

[33] H. Mehdipour, K. Ken Ostrikov, A.E. Rider, Z. Han, Heating and plasma sheath effectsin low-temperature, plasma-assisted growth of carbon nanofibers, Plasma Process.Polym. 8 (2011) 386–400.

[34] H.J. Patrick, A.D. Kersey, F. Bucholtz, Analysis of the response of long period fibergratings to external index of refraction, J. Lightwave Technol. 16 (1998) 1606–1612.

[35] M. Śmietana, W.J. Bock, J. Szmidt, Evolution of optical properties with thickness ofsilicon nitride and diamond-like carbon films deposited by RF PECVD method,Thin Solid Films 519 (2011) 6339–6343.

[36] M. Śmietana, M. Koba, E. Brzozowska, K. Krogulski, J. Nakonieczny, Ł. Wachnicki, P.Mikulic, M. Godlewski, W.J. Bock, Label-free sensitivity of long-period gratings en-hanced by atomic layer deposited TiO2 nano-overlays, Opt. Express 23 (2015)8441–8453.

[37] A.K. Dębowska, M. Śmietana, P. Mikulic, W.J. Bock, High temperature nano-coatedelectric-arc-induced long-period gratings working at dispersion turning point forrefractive index sensing, Jpn. J. Appl. Phys. 53 (2014), 08ME01.

[38] A. Krysiński, M. Śmietana, R. Mroczyński, N. Kwietniewski, W.J. Bock, P. Mikulic,Silicon nitride (SiNx) plasma deposition on optical fiber sensors: coating symmetryperspective, Proc. SPIE 8902 (2013), 89021P.

[39] A. Cusano, A. Iadicicco, P. Pilla, L. Contessa, S. Campopiano, A. Cutolo, M. Giordano,Mode transition in high refractive index coated long period gratings, Opt. Express14 (2006) 19–34.