**1. Introduction**

Supercapacitors are receiving extensive attention worldwide on account of their fast charge/discharge rates, higher power densities, and excellent cycling stabilities, which are superior to other energy storage devices [1–3]. The energy storage mechanisms of supercapacitors are mainly based on electrical double layer capacitance (EDLC), which occurs at the electrode/electrolyte interface [4,5], and pseudo-capacitors with reversible Faradic redox reaction [6]. Their relatively simple energy storage mechanisms and fast charge/discharge traits make supercapacitors the most promising energy storage devices [7–9]. However, the energy that is stored in a supercapacitor is lower than that in batteries. This has inspired research focused on increasing the supercapacitor's energy density, while maintaining high power density [10].

Recently, studies have been conducted to produce porous carbon-based electrode materials, mainly focusing on porous structures [11,12]. In general, specific surface area (SSA) is the most important feature that influences the electrochemical performances of porous carbon materials [1,13,14]. High SSA benefits include good energy storage, especially for high rate performance. On the

other hand, hierarchical pore size distribution accelerates the ion transfer from the electrolyte to electrode surface [15]. Typically, macro- and meso-pores provide space for electrolyte storage, thereby reducing the ion diffusion distance and ion diffusion resistance, which is conducive to better capacitive performance [16–18]. Additionally, three-dimensional (3D) frameworks are capable of supplying stable frame structures, inter-connected pore networks, and are further conducive to ion transfer [18,19]. However, in order to obtain above 3D hierarchical porous carbons (3D-HPCs), it is necessary to add a certain amount of hard templates or soft templates in most of the previously reported works [20–22]. Consequently, the synthesis of 3D-HPCs is limited by its complex, time-consuming, and costly processes.

Surface functional groups also play an important role in energy storage. Nitrogen doping is the most widely studied [23–25]. Generally, a proper amount of nitrogen doping promotes the electrochemical capacitance, by improving the wettability of the porous carbon electrode material, thus bringing about the properties of pseudo-capacitance [26–29]. Nitrogen-doped carbon is achieved by the pyrolysis of nitrogen-enriched polymer precursors and subsequent physicochemical activation, to obtain a nitrogen-doped carbon material. The process is expensive, time-consuming, and endangers our environment [30,31]. On the other hand, through the post-treatment of porous carbon with organic and inorganic nitrogen sources [32], urea [33,34] and ammonia [23,35], it is possible to obtain nitrogen-doped porous carbon. However, the resulting carbon material rarely has the previously mentioned 3D hierarchical structure. The traditional synthesis of 3D hierarchical nitrogen-doped porous carbon (3D-HNPCs) materials is limited to this complex, time-consuming, and costly process. Therefore, transforming sustainable raw materials into highly-performing 3D-HNPC materials, through a simple preparation method, is required.

In this work, we have successfully prepared a 3D hierarchical structure of nitrogen self-doped porous carbon from waste cottonseed husks (CSHs), for high performance supercapacitor electrode materials, using one-step KOH activation. After carbonization and activation, the obtained a-CSHs will have the following features. Firstly, it has a 3D architecture that is associated with hierarchical micro-, meso-, and macro-pores, a high specific surface area of 1694.1 m2/g, and a moderate pore volume of 0.87 cm3/g. Secondly, the obtained a-CSHs also contain a moderate nitrogen content of 2.56 atom %, which is as a result of the protein content in the raw materials. Thirdly, the synthesis process is also simple and convenient for large-scale industrial production. The formation of a 3D hierarchical structure is based on the fact that KOH acts both as hard template to form 3D structures and as an activator to produce affluent micropores on the surface of the carbon material. Finally, cotton is an important agricultural crop in China, with an annual production of 1.5 million tons [36]. The Xinjiang Production and Construction Corps occupies up to 70% of the total production. Cottonseed is mainly used to extract cottonseed oil, thereby generating a large amount of sustainable raw material. On the basis of the above advantages, a-CSHs provide high performed high supercapacitor performances in three- and two-electrode systems.

#### **2. Materials and Methods**

## *2.1. Sample Preparation*

The preparation of nitrogen self-doped three-dimensional (3D) honeycomb-like porous carbon was synthesized by one-step KOH activation. The specific preparation steps were as follows: Cottonseed husk (CSH) was fully washed with deionized water to remove ash and other impurities, and then dried at 100 ◦C for 10 h. It was further crushed to form a powder and passed through a 200 mesh sieve for further use. Subsequently, the CSH powder was vigorously stirred with an aqueous KOH solution at a mass ratio of KOH/CSH powder = 1, and then dried at 80 ◦C. The mixture was then activated in a tube furnace under an Ar atmosphere at 600 ◦C, 700 ◦C, or 800 ◦C for 1 h with a heating rate of 5 ◦C/min. The obtained product was washed with 10% *v*/*v* HCl to remove the metal

impurities, was washed with deionized water until the pH of the filtrate was 7.0, and was dried at 80 ◦C for 10 h. The obtained samples were denoted a-CSH-x, where x represented the activation temperature.

#### *2.2. Material Characterization*

Scanning electron microscopy (SEM) surveys were examined with a Hitachi SU8010 microscope (Tokyo, Japan). Transmission electron microscopy (TEM) and energy dispersive X-ray spectroscopy (EDS) were analyzed by a field emission Tecnai G2 F20 electron (Hillsboro, OR, USA) microscope. X-ray diffraction (XRD) measurements were carried out on a Bruker D8 Advance X-ray diffractometer with Cu-K α radiation (Karlsruhe, Germany). Specific surface areas of the samples were calculated using the Brunauer-Emmett-Teller (BET) method (Micromeritics ASAP 2020 BET apparatus, Atlanta, GA, USA). The pore size distribution (PSD) curves were derived from the adsorption branch, using a nonlocal density functional theory (NLDFT) model assuming slit pore geometry. The surface chemical compositions were determined using an ESCALAB 250Xi (Thermo Fisher Scientific, USA) X-ray photoelectron spectroscope (XPS). The Raman spectra were collected on a LabRAM HR800 Laser Confocal Micro-Raman Spectroscope (Horiba Jobin Yvon, Franch) with a laser wavelength of 532 nm.

#### *2.3. Electrochemical Measurements*

The electrochemical properties of the as-prepared samples were tested on a CHI 760E working station with a 6 M KOH electrolyte. Cyclic voltammetry (CV) tests at different scanning rates and galvanostatic charge/discharge (GCD) curves under varying current densities were used to evaluate the electrochemical performances of the electrode materials. The working electrode was obtained by mixing carbon material (5 mg) with acetylene black (1 mg) and polytetrafluoroethylene (1 μL) in absolute ethanol (1 mL). The mixture was dispersed by ultrasound for 40 min and the ink-like dispersion that was obtained was transferred to nickel foam (1 cm × 1 cm) and then vacuum dried at 80 ◦C for 10 h. The nickel foam was further pressed on a tablet press at 20 MPa for 1 min and was immersed in 6 M KOH for further testing. The loaded mass of each electrode was 5 mg. For the three-electrode system, the Pt sheet and Saturated Calomel Electrode (SCE) were utilized as counter electrode and reference electrode, respectively. The specific capacitances of the samples were calculated through discharge curves following Equation (1), as follows:

$$\mathbf{C} = \frac{I \times \Delta t}{m \times \Delta V} \tag{1}$$

where *C* (F/g) is the specific capacitance, *I* (A) is the charge/discharge current, Δ*t* (s) is the discharging time, *m* (g) is the mass of the working electrode, and Δ *V* (v) is the voltage window of the charge/discharge process.

The electrochemical properties of a-CSH-700 were measured with the two electrode system. Two symmetrical electrodes were separated by a cellulose membrane in a 6 M KOH electrolyte and were assembled in a CR2032 stainless-steel coin cell. The specific capacitance was calculated from the discharge process, according to Equation (1). The energy density and power density of symmetric supercapacitor systems were further calculated by Equations (2) and (3).

$$E = \frac{1}{2} \mathcal{C}\_t \Delta V^2 \times \frac{1}{3.6} \tag{2}$$

$$P = \frac{E}{\Delta t} \times 3600\tag{3}$$

where *E* (Wh/kg), *P* (W/kg), *C*t (F/g), Δ *V* (v), and Δ*t* (h) are the specific energy density, specific power density, specific capacitance, and voltage window, respectively, of the symmetrical supercapacitor system.
