Tensor is valid for a single CNC crystal, but for afilm containing randomly aligned crystals (such asCNF film), the overall piezoelectric response is acombination of responses of all crystals. Due tothis, maximal piezoelectric response would resultfrom perfectly aligned structure as below:

A water suspension of bleached sulphite birch fibers, processed through a Masuko grinderusing three consecutive passes and further homogenized using six passes at 2000 barpressure by using a M110P microfluidizer equipped with 200 and 100 µm chambers [8].The resulting CNF material is composed of crystalline (cellulose I) [1] and amorphousdomains. A self-standing CNF films were fabricated by pressure filtering (15-30 min)followed by pressing and drying in a hot-press at 100 °C for 2 h [9].

Nanocellulose aerogels can be produced by free-drying method, where the aqueousnanocellulose dispersion is frozen and subsequently dried in vacuum by sublimation,leaving a highly porous cellulose structure, also known as nanocellulose aerogel.

We report the experimental results on piezoelectricity of wood-based and bacterialnanocellulose films. Films of wood-based cellulose nanofibrils (CNF) and cellulosenanocrystals (CNC), and bacterial nanocellulose (BNC) show piezoelectric sensitivityvalues that align between the piezoelectric coefficients of quartz (2.3 pC/N) andpolyvinylidenefluoride (PVDF, -33 pC/N). Orientation/polarization is expected to increase thepiezoelectric sensitivity of nanocellulose films. Also nanocellulose aerogels embedded intoan elastic polydimethylsiloxane (PDMS) films showed piezoelectric response. Our resultssuggest that nanocellulose is a suitable precursor for disposable piezoelectric sensorswith potential applications in the fields of electronics, sensors and biomedical diagnostics.

Sampo Tuukkanen, Satu Rajala, Marlitt Viehrig and Pasi KallioTampere University of Technology (TUT), Department of Automation Science and Engineering, P.O. Box 692, FI-33101 Tampere, Finland.

[1] R. J. Moon, A. Martini, J. Nairn, J. Simonsen, and J. Youngblood, Chemical Society Reviews 40(7), 3941 (2011).[2] S. Tuukkanen, S. Lehtimäki, F. Jahangir, A.-P. Eskelinen, D. Lupo and S. Franssila, “Printable and disposable supercapacitor from nanocellulose and carbonnanotubes”, Proceedings of Electronics System-Integration Technology Conference (ESTC), (2014).[3] K. Torvinen, S. Lehtimäki, J. T. Keränen, J. Sievänen, J. Vartiainen, E. Hellén, D. Lupo and S. Tuukkanen, “Pigment-Cellulose Nanofibril Composite and Its Applicationas A Separator-Substrate in Printed Supercapacitors” Electronic Materials Letters 11(6), 1040 (2015).[4] E. Fukada, Journal of the Physical Society of Japan 10(2), 149 (1955).[5] L. Csoka, I. C. Hoeger, O. J. Rojas, I. Peszlen, J. J. Pawlak and P. N. Peralta, ACS Macro Letters 1(7), 867 (2012).[6] B. Frka-Petesic, B. Jean and L. Heux, Europhysics Letters 107(2), 28006 (2014).[7] S. Rajala, M. Mettänen and S. Tuukkanen, “Structural and Electrical Characterization of Solution-Processed Electrodes for Piezoelectric Polymer Film Sensors”, IEEESensors Journal 16(6), 1692 (2016).[8] M. Pääkko, M. Ankerfors, H. Kosonen, A.; Nykänen, S. Ahola, M. Österberg, J. Ruokolainen, J. Laine, P. T. Larsson, O. Ikkala, T. Lindström, “Enzymatic HydrolysisCombined with Mechanical Shearing and High-Pressure Homogenization for Nanoscale Cellulose Fibrils and Strong Gels”, Biomacromolecules 8(6), 1934 (2007).[9] M. Österberg, J. Vartiainen, J. Lucenius, U. Hippi, J. Seppälä, R. Serimaa, J. Laine, “A Fast Method to Produce Strong NFC Films as a Platform for Barrier andFunctional Materials”, ACS Appl. Mater. Interfaces 5(11), 4640 (2013).[10] S. Rajala, M. Vuoriluoto, O. J. Rojas, S. Franssila and S. Tuukkanen, “Piezoelectric sensitivity measurements of cellulose nanofibril sensors”, Proceedings of IMEKO2015 World Congress “Measurement in Research and Industry.[11] S. Tuukkanen and S. Rajala, “A survey of printable piezoelectric sensors”, IEEE Sensors 2015 Conference Proceedings.[12] Satu Rajala, Tuomo Siponkoski, Essi Sarlin, Marja Mettänen, Maija Vuoriluoto, Arno Pammo, Jari Juuti, Orlando J. Rojas, Sami Franssila and Sampo Tuukkanen,“Cellulose nanofibril film as a piezoelectric sensor material”, (submitted).

Cellulose based nanomaterials, generally known as nanocellulose, can be fabricated from woodcellulose fibers [1]. Cellulose nanofibrils (CNF), produced by a mechanical homogenizing processfrom cellulose fibers, contain both crystalline and amorphous regions. Cellulose nanocrystals(CNC) can be obtained from CNF by removal of amorphous regions using hydrolysis e.g. insulfuric acid. A strong ability to form light-weight, highly porous, entangled networks makesnanocellulose e.g. suitable substrate and/or separator for supercapacitors [2, 3].

Chemical structure of cellulose:

Cellulose crystal unit cell:

The compression of a piezoelectric film under an external force causes charge separationgenerating a voltage between sensor sides. The piezoelectric coefficient dmn defines the chargedensity generated by a certain applied stress. Wood piezoelectricity resulting from crystallineassemblies of cellulose was proposed already in 1950’s [4], but experimental evidence ofpiezoelectricity of CNC was reported only very recently [5, 6]. The piezoelectric tensor dmn forwood is determined by the symmetry of a crystal lattice. The monoclinic C2 symmetry and thecancellation effects result into following matrix, where d14 = -d25 [4]:

CNCwhiskers

Amorphousnanocelluloseor othermedium

Ø In-house built setup equipped with a shaker (Brüel & Kjaer Mini-Shaker), reference sensors fordynamic and static forces, and a charge amplifier [7]

Ø Measured sensitivity value is closely related to transverse piezoelectric coefficient d33

Ø A static force of ~3 N keeps a sensor sample in place. Excitation with sinusoidal 2 Hz inputresults in a dynamic force of ~1.3 N. Typically 6-18 excitations per sensor are done. The sensorsensitivity (in pC/N) is given by generated charge divided by the dynamic force.

Certain bacterial strains (Acetobacter xylinum and Agrobacterium tumefaciens) can begenetically engineered to produce large quantities of pure cellulose from e.g. glucose andsugar. In laboratory conditions, a bacterial cellulose film can be produced by cultivating asuitable strain of bacteria in steady growth medium with a large amount of sugar.

7.0 µm

(The aerogels fabricated by Saeed Karimi at TUT. The µCT images by Ilmari Tamminen from Porf. Jari Hyttinen’s group at TUT’s BioMediTech.)

1 mm

÷÷÷

ø

ö

ççç

è

æ=

0000000000000000

25

14

dd

dmn

Ø Piezoelectric coefficients matrix dmn (3 × 6)Ø Electrical axis m = 1, 2, 3 & mechanical axis n = 1, 2,…, 6Ø Length = 1, width = 2 and thickness = 3Ø Shear stress around axes is represented by 4, 5 and 6.

[C6H10O5]n

(SEM images by Essi Sarlin from TUT’s Dept. Material Science)48 µm

(BNC film fabricated by Arno Pammoat Prof. Matti Karp’s laboratory at TUT)

500 nm500 nm 1 µm1 µm

CNC aerogelCNF aerogel

CNC aerogel 1

BNC-3.6um 4

BNC-3.6um 3

BNC-7um 2

BNC-7um 1

CNC T2 [11]

CNC T1 [11]

CNF T2 [11]

CNF T1 [11]

CNF 5 [12]

CNF 4 [12]

CNF 3 [12]

CNF 2 [12]

CNF 1 [12]

0 2 4 6 8 10 12Sensitivity (pC/N)

Nan

ocel

lulo

sepi

ezo-

sens

ors

Total side-1 side-2

5 µm 5 µm

PDMS was used as a stretchable and transparentnanocellulose embedding medium. Aerogels were slowlyimmersed into liquid PDMS followed by removal of trappedair cavities inside the aerogel through a vacuum and curingfor 10h at 60°C until solid. CNF aerogel-PDMS compositesshowed more air entrapment and challenging immersion thenCNC aerogel-PDMS composites. Assembled CNC-PDMScomposite sensors showed sensitivities of 2.6–12 pC/N,yielding 0.006–0.017 (pC/N)/µm for 0.15–2 mm thick films,only slightly higher than pure PDMS giving ~0.004(pC/N)/µm. For thinner films, composite response wasrelatively higher, which may be due to aerogel dimensions.Aerogel cavities and fiber orientation diversity may causehigh standard deviations observed for the composite sensors.

The piezoelectric sensor elements were assembled by sandwiching the CNF filmbetween two copper electrodes evaporated on polyethylene terephthalate (PET)substrate using adhesive film. Measured CNF sensor sensitivities were 4.7–6.5 pC/N [12],yielding 0.10–0.14 (pC/N)/µm for 48 µm thick film. BNC sensors showed sensitivities of2.8–7.8 pC/N, yielding 0.64–0.88 (pC/N)/µm for 3.6 and 7.0 µm thick films. Linearitymeasurement (charge vs. force curve) were performed for CNF sensor and PVDFreference sensor.

PETCu

CuPET

Nanocellulose film

125 µm

100 nm

100 nm125 µm

CNF: 45 µm orBNC: 3.6 or 7 µm

7.0 µmthick

3.6 µmthick