Synthesis of cellulose nanocrystals: Cellulose nanocrystals (CNCs) were synthesized following the protocol described in two previous publications.[1,2] 10 g of microcrystalline cellulose (MCC) (Sigma-Aldrich, ref 310697) were dispersed in 45 ml of ultrapure water using an ultrasound bath (45 kHz) for 5 min. Then, 45 ml of H2SO4 98 wt% (Labkem, ref SUAC-00A) were added dropwise to the mixture (externally cooled in an ice/water bath at 0 °C) reaching a final H2SO4 concentration of 64 wt%. This addition was performed fast (less than 5 min) and with magnetic stirring to obtain type I CNCs or slow (longer than 100 min) and with high shear mixing to obtain type II CNCs.[1] After that, the mixture was put in a heating plate at 70°C for 10 minutes for type I CNCs or at ambient temperature for 1 hour for type II CNCs. The reaction was stopped by diluting in 1 liter of ultrapure water at 4°C, also to increase the pH, and the mixture was left decanting overnight at 4°C. The bottom sediment was dialyzed against ultrapure water until a neutral pH was achieved, using specific dialysis membranes (Merck, average flat width 33 mm, ref D9652-100FT). The neutralized dispersions were subjected to centrifugation/re-dispersion cycles at 9000 rpm (9327 rcf) for 1 min to keep only the nanocrystals. The average mass yield and the concentration (Table S1) were determined by lyophilization of 3 aliquots of 30 mL of the CNCs dispersions. The hydrodynamic radii, polydispersity, and ζ-potential were measured using a Malvern Nano ZS instrument, and the mean values of these parameters are presented in Table S1. The CNCs in solid were measured by a Bruker D8 Advance X-Ray diffractometer using a Cu tube as the X-ray source (λ Cu Kα = 1.54 Å), a tube voltage of 40 kV, and a current of 40 mA. The X-Ray diffraction patterns of the prepared CNCs (Fig. S1) confirmed the different presence of the characteristic planes of type I or type II cellulose.--

Synthesis of cellulose nanocrystals: Table S1. Mean values of the characterization results of type I and type II CNCs dispersions. Fig. S1. X-ray diffraction profiles of type I (red) and type II (blue) CNCs. Surface resistivity of n-type thermoelectric cotton textiles as a function of the number of dip-coating cycles: Fig. S2. Surface resistivity (Rsh) of nanocomposite textiles from 1 to 10 immersion cycles with equivalent inks, but different immersion methodology: regular (red) and using a complementary ultrasounds bath treatment for 5 min (black). Measured with an in-line 4-point probe configuration. Error bars show standard deviation of at least 3 repetitions. SEM images of the surface of original and washed CWF@CNF textile samples: Fig. S3. SEM images of the surface of original and washed CWF@CNF textile samples at different magnifications (a) CWF@CNF, (b) CWF@CNF W30 and (c) CWF@CNF W45. Model proposed for describing the nonlinear Seebeck of CNFs and CWF@CNF: The model presented in this work represents the combination of two physical mechanisms occurring in parallel: S(T)=S_met (T)+S_imp (T) (1). Power factor and estimative figure of merit of unwashed and washed samples and CNFs: Table S2. Thermoelectrical properties of dip-coated textiles and carbon nanofibers. References.

Financial support from Spanish MCIN/AEI under projects PID2019-104272RB-C51/AEI/10.13039/501100011033 and PID2020-120439RA-I00, from Spanish CSIC (PIE iniciación ref. 202280I007), as well as from Gobierno de Aragón (DGA, Grupo Reconocido DGA-T03_23R) is acknowledged. Vícto Calvo thanks the DGA for funding his PhD contract (Ref. CUS/581/2020). Antonio J. Paleo gratefully acknowledges support from FCT-Foundation for Science and Technology by the “plurianual” 2020–2023 Project UIDB/00264/2020, and European COST Action EsSENce CA19118 for its support with the Short Term Scientific Mission (STSM) grant E-COST-GRANT-CA19118-0ed3a197 at IPF (Dresden). E. Muñoz acknowledges funding from ANID ANILLO ACT/192023 and Fondecyt No 1230440.

Under a Creative Commons license BY-NC-ND 4.0 Deed.

Peer reviewed