Thickness and substrate dependences of phase transition, drug release and antibacterial properties of PNIPAm-co-AAc films

Abstract

Micron and submicron poly(N-isopropylacrylamide)-co-(acrylic acid) (PNIPAm-co-AAc) films were deposited onto silicon and gold substrates by the spin-coating procedure. The influence of polymer–substrate interaction and spatial confinement of macromolecular chains in the ultrathin polymer films on lower critical solution temperature (LCST) was investigated under different pH conditions. Shift and broadening of the LCST temperature range was observed from the critical thickness of polymer film. It was also found that the substrate plays a key role in this shift. The observed phenomenon was applied for the temperature-controllable release of a small molecular dopant (crystal violet, CV) from the ultrathin polymer films. Finally, doped ultrathin polymer films were examined for their antibacterial activity by in-contact and drop methods. It was observed that polymer thickness and support substrate can influence both CV release and antibacterial properties. Despite the fact that the concentration of CV used in PNIPAm-co-AAc was constant and thinner films contained a significantly smaller amount of CV than thicker ones, the antibacterial activity of thin films was found to be greater in several cases.

---

Introduction

Stimuli-responsive polymers are defined as materials that sense and respond to changes in their environment by changing their physical or chemical state. The stimulus can be a change of pH and temperature,¹ introduction of specific analytes,² electric and magnetic fields or other external stimuli.^3,4 Responsive materials reacting by roughness, wettability, biocompatibility, and optical appearance modulation find applications in sensing and delivery of functional molecules.^5,6 Developing stimuli-responsive systems for controlled and tunable drug release is of a particular interest.^7,8

One of the most studied stimuli-responsive polymers is PNIPAm, which undergoes a volume phase transition near 31 °C (LCST – lower critical solution temperature). PNIPAm is fully water-soluble below the LCST and loses its solubility above this temperature.⁹ From the practical point of view it is sometimes necessary to tune the LCST into a different temperature range. A basic way to shift LCST value is copolymerization with other monomers. Depending on the used monomers LCST can be shifted by several tens of Celsius degrees¹⁰ and, moreover, some additional sensitivity can be gained.¹¹ The most common comonomer is acrylic acid (AAc), which adds responsivity to acidity. At pH > 4.25 the AAc groups are deprotonated and the polymer swells. When the pH decreases under 4.25 again the polymer returns back to its initial state. Due to this thermal and pH responsivity, the PNIPAm-co-AAc can find application as a special sensor component, drug delivery system and antifouling coating.^12–14

The use of PNIPAm-co-AAc in a thin film format will combine its smart properties with an arrangement that can be more easily influenced by surrounding environment. It is well known that spatial confinement in ultrathin polymer films can significantly change their properties.^15,16 The two main phenomena which affect the ultrathin film behavior are: (i) spatial confinement and (ii) interface and surface effects.^15,17 The influence of substrate in regard to the interface effect was studied in several works.^14,18 However, the observed dependences were rather related to PNIPAm-co-AAc density and arise mainly during the deposition procedure. Moreover, attention was dominantly paid only to the interaction of polymer with gold substrate, which is important in the preparation of the so-called “ethalon” component, in which PNIPAm layer is sandwiched between two Au layers.¹⁴

Confinement effect was also reported to influence the polymer phase transition and the release of small molecules from thin films. Like the previous case most of the published studies are related to bilayer or “sandwiched” systems, where the responsive polymer is coated by a layer of another material.^19,20 Release of the incorporated compound was determined by the thickness of the cladding layer²¹ or the polymer-cladding interaction.¹⁹

This study was inspired by these efforts and by the previous studies of Cory et al.^15,17 We have investigated the changes of properties related to spatial confinements of macromolecular chains in ultrathin polymer films and polymer/substrate interaction, as well as a potential tuning of pH and temperature responsivity by the abovementioned phenomena. To obtain detailed information of the thickness dependent LCST tuning we have utilized in situ reflection spectroscopy. The observed shift of LCST was employed for controllable release of antibacterial agent from ultrathin PNIPAm-co-AAc polymer film.

Experimental

Materials

Poly(N-isopropylacrylamide)-co-(acrylic acid) (PNIPAm-co-Aac, acrylic acid 15 mol%) and crystal violet (tris(p-dimethylaminophenyl)methyl chloride, CV) were obtained from Sigma Aldrich and used without further purification. Chloroform p. a. was purchased from LachNer. Mueller-Hinton agar (MHA) was prepared as described by producer (Oxoid, CM0337) and sterilized in autoclave.

Samples preparation

PNIPAm-co-AAc was dissolved in chloroform (0.1–10 wt%, depending on the desired film thickness) and deposited by the spin-coating technique at 1000 rpm on silicon (Si) or gold (Au) substrates. Gold substrates (on glass) were prepared by vacuum sputtering on Balzers SCD 050 device from gold targets (99.95%, supplied by Goodfellow, Ltd). The deposition conditions were: DC Ar plasma, gas purity 99.995%, room temperature (25 °C), sputtering time 100 s, deposition current of 20 mA at Ar pressure of 4 Pa. For doping of polymer with CV the compounds were separately dissolved in chloroform and mixed together with constant weight rate (PNIPAm-co-AAc/CV = 99/1). Optical clarity and absence of light scattering indicate the CV is uniformly distributed and forms a molecular dispersion in the PNIPAm-co-AAc matrix.

Measurement techniques

The film thickness was measured (standard error ± 10%) by a profilometer Hommel 1000 and atomic force microscopy (AFM) in contact mode by scratch method. The surface morphology was examined using AFM in tapping mode under ambient conditions on a Digital Instruments CP II set-up. Veeco oxide-sharpened silicon probes RTESPA-CP with the spring constant of 40 N m⁻¹ were used. Mean roughness (R_a) was calculated as arithmetic average of the deviations from the center plane of the sample.

Refraction spectra of thin film deposited onto Si and Au substrates were determined in situ under cooling from 60 °C to 25 °C in 250–750 nm spectral range using refractometer Avaspec 2048. After each set of measurements the wavelength, at which the change of refraction intensity due to polymer phase transition is maximal, was determined and used as a characteristic.

Crystal violet release from polymer films was determined using absorption UV-Vis spectroscopy. Samples were immersed in distilled water for different time intervals at 24 °C and 40 °C. Absorbance of extracted solution was measured using spectrometer Lambda 25 (PerkinElmer, Inc.) in 250–800 nm wavelength range. After background correction the absorption coefficient on the 590 nm wavelength (maximum of CV absorption) was determined and used as a parameter characteristic for CV release.

Antibacterial tests

The antibacterial effect of PNIPAm-co-AAc films doped with CV was carried out using Gram-positive Staphylococcus epidermidis strain from the Microbial Type Culture Collection (University of Chemistry and Technology, Prague). Antibacterial effectivity was measured using two approaches: after the extraction of CV (drop method) or contact method. In the first case doped polymer films were immersed in 2 mL of buffer solution (pH = 4) for 3 h under 25 °C and 40 °C. Overnight culture of S. epidermidis derived from a single colony and cultivated in Luria–Bertani broth medium were used for the experiment. The sample extracts were neutralized by alkaline buffer (pH = 8.5) and 0.5 mL of resulted solution was inoculated with the 0.5 mL of bacterial suspension (final concentration of 1000 mL⁻¹ colony forming unit, CFU). Bacterial samples incubated only in the pristine physiological solutions served as controls. The inoculated solutions were incubated at laboratory temperature in static conditions for 24 h and aliquots of 25 μL from all samples were placed on LB agar plates in multiple replicates. The growth of S. epidermidis was evaluated after 24 h of growth at 37 °C. Each sample was prepared separately in a triplicate.

Alternatively, the samples were attached onto agar plates previously colonized by bacteria (1 mL of bacteria suspension containing 5107 of S. epidermidis cells) and kept in contact for 24 h at 37 °C. As control a PNIPAm-co-AAc thin films without CV were used. Evaluations of inhibition zone appearance and development were performed after 20, 22, and 24 h of samples placement in contact with bacterial strain.

Results and discussion

The chemical structures of the stimuli responsive polymer (PNIPAm-co-AAc) and the antimicrobial agent (CV) are represented in Fig. 1. The PNIPAm-co-AAc was chosed due to its pH and temperature sensitivity and CV is a well known antimicrobial agent, which has an additional beneficial feature – a strong absorption band in the UV-Vis optical range.

Fig. 1 Chemical structures of: A – antimicrobial Crystal Violet (CV); B – stimuli-responsive PNIPAm-co-AAc polymer.

Relative reflection of light (at 450 nm) from 1 μm thick PNIPAm-co-AAc films deposited on the Si substrate is presented in the Fig. 2 for neutral and acidic environment. Polymer films were previously heated, immersed into distilled water and gradually cooled. Used method allows continuous and noninvasive measurement of polymer phase state transition that manifests itself as apparent change of reflection. From measured data it is possible to estimate both temperature and temperature range of phase transition. For clarity, temperature range of phase transition at pH = 4 is depicted by dotted lines. It is obvious, that at the neutral pH phase transition occurs in the 30–32 °C temperature range. With pH decreasing the phase transition tends to occur at higher values of temperature. It can be expected that the decreasing of pH leads to protonation of AAc and PNIPAm-co-AAc becomes less soluble. In all cases phase transition shows sharp profile and occurs in 1.5–2 °C wide range.

Fig. 2 Temperature dependence of reflection coefficient (by 450 nm) from 1 μm thick PNIPAm-co-AAc films deposited on Si substrate at different pH. Dotted lines indicate the initial (t₁), center (t_c) and final (t₂) points of PNIPAm-co-AAc phase transition at pH = 4.

Measurements similar to those presented in Fig. 3 were performed on micron and submicron thick PNIPAm-co-AAc films deposited onto Si and Au substrates. Values of the central temperature and the temperature range of PNIPAm-co-AAc phase transition were obtained and presented in Fig. 3 as a large symbol (central point of phase transition) and two small symbols (beginning and end of the phase transition). It is apparent that with decreasing thickness of polymer, phase transitions are shifted to higher values. This shift becomes more pronounced for polymer film with thickness below 200 nm. Simultaneously, the temperature range broadens significantly. This phenomenon is more pronounced in the case of PNIPAm-co-AAc films deposited onto gold substrates. Environmental pH also affects the thickness dependences – under acidic pH the observed phenomena are more significant.

Fig. 3 Temperature dependence of phase transition at pH = 4 and pH = 7 of PNIPAm-co-AAc film deposited on Au and Si substrate.

Surface morphologies of pristine polymer films and films soaked at 50 °C are shown in Fig. 4. AFM scans provide information about changes of a thin area close to surface of PNIPAm-co-AAc. Polymer surface after preparation is rough, which can be attributed to the deposition – spin-coating is quick, nonequilibrium process, where the macromolecular chains are rapidly frozen, leading to an increased roughness in the nanoscale. During the soaking procedure polymer surface becomes smoother. It can be concluded that the water penetrates into the surface layer of the polymer. This process increases the free volume and unfreezes the polymer chains, leading to transition into more favorable conformation and surface smoothening. So, despite the fact the PNIPAm-co-AAc is completely insoluble at pH = 4 and temperatures above LCST, water can penetrate into the polymer film. The amount of water absorbed this way is certainly significantly lower than at pH > 7 or temperatures below LCST.

Fig. 4 Surface morphology of 50 nm and 500 nm thick PNIPAm-co-AAc polymer films before and after soaking at pH = 4 and 25 °C for 3 hours.

Observed thickness dependences were applied to control the release of CV from thin and ultrathin PNIPAm-co-AAc films deposited on Si substrates. (The release of CV from Au is presented in ESI.† In this case released curves do not show as good reproducible, as in the case of Si substrate, probably during to Au surface inhomogeneities, arising during sputtering procedure.) Results of these measurements are presented in Fig. 5. CV was extracted by distilled water and optical absorption coefficient of extracts was measured at 590 nm. It must be noted that the weight concentration of CV in all PNIPAm-co-AAc films was constant, so 500 nm thick polymer films contain ten times the amount of CV as 50 nm thick films. Thus instead of the absolute value of the absorption the thickness normalized value is presented. This variable can be physically interpreted as the amount of CV extracted from a unit thickness. Below LCST (at 25 °C), both “thin and thick” (50 and 500 nm) polymer films are partially soluble. Thus, swelling of the polymer occurs accompanied by CV release. According to Fig. 3, 50 nm thick PNIPAm-co-AAc film have transition temperature of 44 °C at pH = 7 and 52 °C at pH = 4. On the other hand, the transition temperature of thicker polymer film is similar to the “bulk” value (near 32 °C). So at 25 °C thin polymer film is further from its LCST than the thick one. This explains the greater released amount of CV from 500 nm thick polymer film. On the other hand, at temperatures under 40 °C thick polymer film is well above its LCST. Under these conditions PNIPAm-co-AAc is fully insoluble, so the release of CV is restricted. The thin polymer film is not above its LCST and the release of CV occurs and is similar to the case of 25 °C. Amounts of leached CV were calculated from the Fig. 5 using the calibration curves obtained by gradual dilution of CV solution with known concentration. Obtained values are presented in Table 1. pH-dependent differences of the CV release can be explained by PNIPAm-co-AAc deprotonation. Polymer with acidic functional groups such as –COOH deprotonates at basic pH and acquires a negative charge. Due to the high amount of similarly charged functional groups polymer chains repel each other and greater amount of water can penetrate in the polymer, allowing the quick and complete release of CV from polymer film. The Fick diffusion can also contribute to a faster release of the CV from the thinner films. Apparently, to be released from thin PNIPAm-co-AAc film CV molecule needs to pass shorter distance than from thick film. The data on release of CV from Au substrates is presented in ESI.† In this case, the curves of CV release were not so well reproducible as in the case of Si substrate, which was probably caused by Au surface inhomogeneities arising from the sputtering procedure.

Fig. 5 Normalized absorption coefficient at 590 nm of solution extracted CV from 50 nm and 500 nm thick PNIPAm-co-AAc films deposited on Si substrates. Releases were measured for temperature below (25 °C) and above (40 °C) LCST.

Table 1 Relative amounts of leached CV (%) from 50 and 500 nm thick polymer films deposited onto Si substrate at different temperature and pH

t [°C]	pH	Thickness 50 nm	Thickness 500 nm
25	7	91 ± 3.3	62 ± 10.7
	4	58 ± 1.5	44 ± 16.3
40	7	49 ± 13.2	11 ± 6.2
	4	38 ± 13.9	7 ± 2.0

---

Finally, the primary antibacterial activity of CV doped polymer films was verified using a Gram-positive bacteria strain S. epidermidis. It must be noted that CV was chosen not only because of its strong absorption peak at 590 nm for release measurement, but also due to its activity against Gram-positive bacteria. Samples of the solutions were obtained after 3 hours of CV extraction at constant temperatures (25 °C and 40 °C) and pH = 4 and pH = 7. 50 and 500 nm thick polymer films deposited on the Si and Au substrates were tested. After sampling, the samples were in contact with bacteria overnight and observations of bacterial growth on agar were performed after 24 h. The results of the antibacterial tests are presented in Fig. 6. Control measurements of bacterial growth were performed: (i) in suspension diluted by distilled water (bacterial control), (ii) in suspension diluted by extracted solutions without CV (solution control), and (iii) in solution with known CV concentration (concentration which would be obtained by complete dissolution of the film). It is evident that the extracts obtained at pH = 7 show a considerable antimicrobial activity due to the “complete” release of CV from partially deprotonated and swollen polymer. In case of CV release at pH = 4, the antimicrobial activity is strongly dependent on the samples thickness as well as on extraction temperature. From Fig. 6 it is evident that the number of colony forming unit (CFU) decreased with increasing dilution of the extracted solution. This can be attributed to the high solvent strength of the solution, which can eliminate bacterial growth. Further significant decrease of CFU occurs in the case of doped samples due to CV leaching from polymer films. It is apparent that in the case of CV extracted from thick films antibacterial activity depends strongly on the temperature: solution soaked at 25 °C exhibits greater antibacterial activity than the one soaked at 40 °C. At the extraction temperature of 25 °C both thick and thin polymer films are below LCST. Antimicrobial activities are therefore determined by the total amount of CV that can be extracted from the polymer films, which is higher in the case of thicker films (Si500 nm and Au500 nm). This result can be expected, because at 40 °C thick polymer films are above LCST, do not swell and CV release is restricted. Situation is inversed in the case of thin polymer films. Thin polymer films remain below their LCST at 40 °C and the CV extraction is more effective at the higher temperature, leading to greater antibacterial activity of the CV solution obtained at this temperature. Observed results correspond well with Fig. 5. It is also evident that substrates affect antibacterial properties of thin polymer films. Solutions from films deposited onto Au substrates are more effective. These results correspond with Fig. 3, where the Au substrate promotes the increase (and widening) of LCST, allowing the film to get more swelled. Regarding the comparison of CV release tests (Fig. 5) and results of antimicrobial activity two key points must be taken into account: release from thin film is quicker and much higher amount of CV is released than from the thick films. On the other hand, thick film contains considerably more CV (concentration of CV is constant while the thickness is ten times higher). It is evident the first point (thickness effect and related quick and full release) is more critical in the case of antimicrobial test performed on extracts, taken at 25 °C, where the thin films exhibit significantly greater antimicrobial activity. On the other hand the situation is inversed in the case of antimicrobial extracts performed at 40 °C. Probably, at this temperature Fick diffusion from the bulk of film becomes easier and as a result the amount of incorporated CV becomes the decisive factor, affecting the antimicrobial properties. Results of the antimicrobial tests correspond well with the relative amounts of leached CV (Table 1). Release at 25 °C, when both thick and thin films are below LCST, led to more pronounced antimicrobial activity of extracts from thick films due to greater amount of incorporated CV. In opposite, release at 40 °C, where the thick film is above LCST unlike the thin film, led to greater antimicrobial activity of a thin film, because of faster and complete CV release.

Fig. 6 Number of colony forming units (CFU) on samples after treatment by solutions extracted from 50 and 500 nm thick PNIPAm-co-AAc films with CV on Au and Si substrates at 25 and 40 °C in comparison with control samples.

Results of antibacterial tests performed in contact arrangement are presented in Table 2. Similarly to the case of Fig. 5 the inhibition zone sizes were as well normalized for determination of antibacterial effect per thickness unit. Samples were in contact with bacteria at the temperature of 37 °C. This temperature is above the phase transition temperature of the thick polymer, but below the transition temperature of the thin one. Thus it was initially expected that thick polymer film will not create a zone of inhibition. However, in this case inhibition zones were well visible. The release of CV probably occurs above LCST due to particular swelling of the surface layer (Fig. 4 and 5). The Table 2 illustrates that with the decrease of polymer thickness the normalized size of inhibition zone increases. Influence of the Si and Au substrate was found to be less significant compared to the case of Fig. 6. On the base of the in-contact antibacterial tests supported by the solution experiments it can be concluded that the antibacterial activity of CV doped PNIPAm-co-AAc films increases with decreasing thickness. This effect can be attributed to the shift of PNIPAm-co-AAc phase transition from “critical” thickness.

Table 2 Dependences of the size of normed inhibition zone (in cm) on the thickness of PNIPAm-co-AAc films doped with crystal violet spin-coated on Si or Au substrate

Si
Time of bacteria growth [h]	20			22			24
Thickness [μm]	1.0	0.5	0.05	1.0	0.5	0.05	1.0	0.5	0.05
Normed inhibition zone size (−)	7.6	9.6	78	5.3	6.4	46	5.8	7	44

---

Au
Thickness [μm]	1.0	0.5	0.05	1.0	0.5	0.05	1.0	0.5	0.05
Normed inhibition zone size (−)	6.6	9.4	62	6.1	8	58	5.5	5.8	50

---

Conclusion

In this work the thickness and substrate influence on the LCST in thin polymer films deposited by spin-coating are reported. It was found that decrease of polymer thickness below a critical value leads to a shift of the LCST to a higher value. This effect also depends on the substrate. In the case of Au substrate shift was found to be more pronounced. Observed phenomenon was applied to tune the release properties of PNIPAm-co-AAc doped with CV. Significant differences in the release per unit thickness was observed in the case of thin and thick polymer films. Finally, antibacterial properties of doped films against Gram-positive bacteria were studied using drop (CV independent extraction) and contact (direct contact between doped polymer and bacteria strain) methods. It was found that antimicrobial activity per thickness unit of polymer film significantly increases with decreasing thickness of the polymer. This phenomenon can be explained by the phase shift of LCST and easier leaching of CV from thin PNIPAm-co-AAc layer.

Acknowledgements

This work was supported by the GACR under the projects 15-19485S and by the Ministry of Health of CR under the project 15-33459A.

References

1. D. Schmaljohann, Adv. Drug Delivery Rev., 2006, 58, 1655–1670.
2. R. V. Ulijn, N. Bibi, V. Jayawarna, P. D. Thornton, S. J. Todd, R. J. Mart, A. M. Smith and J. E. Gough, Mater. Today, 2007, 10, 40–48.
3. D. Roy, J. N. Cambre and B. S. Sumerlin, Prog. Polym. Sci., 2010, 35, 278–301.
4. O. Lyutakov, O. Hejna, A. Solovyev, Y. Kalachyova and V. Svorcik, RSC Adv., 2014, 4, 50624–50630.
5. Q. M. Zhang, W. Xu and M. J. Serpe, Angew. Chem., Int. Ed., 2014, 53, 4827–4831.
6. M. A. C. Stuart, W. T. S. Huck, J. Genzer, M. Muller, C. Ober, M. Stamm, G. B. Sukhorukov, I. Szleifer, V. V. Tsukruk, M. Urban, F. Winnik, S. Zauscher, I. Luzinov and S. Minko, Nat. Mater., 2010, 9, 101–113.
7. A. K. Bajpai, S. K. Shukla, S. Bhanu and S. Kankane, Prog. Polym. Sci., 2008, 33, 1088–1118.
8. R. Elashnikov, O. Lyutakov, Y. Kalachyova, A. Solovyev and V. Svorcik, React. Funct. Polym., 2015, 93, 163–169.
9. S. Minko, Handbook of Stimuli-Responsive Materials, ed. M. W. Urban, Wiley-VCH, Published online, 2011.
10. I. C. Barker, J. M. G. Cowie, T. N. Huckerby, D. A. Shaw, I. Soutar and L. Swanson, Macromolecules, 2003, 36, 7765–7770.
11. J. Zhang, L. Y. Chu, Y. K. Li and Y. M. Lee, Polymer, 2007, 48, 1718–1728.
12. K. C. C. Johnson, F. Mendez and M. J. Serpe, Anal. Chim. Acta, 2012, 739, 83–88.
13. D. Parasuraman and M. J. Serpe, ACS Appl. Mater. Interfaces, 2011, 3, 4714–4721.
14. H. Huang and M. J. Serpe, J. Appl. Polym. Sci., 2015, 132, 42106.
15. J. L. Keddie, R. A. L. Jones and R. A. Cory, Europhys. Lett., 1994, 27, 59–64.
16. V. Svorcik, O. Lyutakov and I. Huttel, J. Mater. Sci.: Mater. Electron., 2008, 19, 363–367.
17. J. L. Keddie, R. A. L. Jones and R. A. Cory, Faraday Discuss., 1994, 98, 219–230.
18. O. Zavgorodnya and M. J. Serpe, Colloid Polym. Sci., 2011, 289, 591–602.
19. Y. Gao, G. P. Zago, Z. Jia and M. J. Serpe, ACS Appl. Mater. Interfaces, 2013, 5, 9803–9808.
20. K. Vasilev, N. Poulter, P. Martinek and H. J. Griesser, ACS Appl. Mater. Interfaces, 2011, 3, 4831–4836.
21. A. Cavallaroa and K. Vasilev, Chem. Commun., 2015, 51, 1838–1841.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra13972d

---