Thermally reduced graphene oxide as an electrode for CDI processes: A compromise between performance and scalability?

Ayelet Kalfa, Tirupathi Rao Penki, Izaak Cohen, Netanel Shpigel, Eran Avraham, Doron Aurbach, Dawei Liang, Qinghao Wu, Haining Wang, Yan Xiang

Research output: Contribution to journalArticlepeer-review

10 Scopus citations


Capacitive deionization (CDI) is an alternative water desalination technology, which was investigated extensively in the last decade. The choice of electrodes' materials plays a major role in the electrosorption performance, affecting the whole desalination process. Graphene-based nanostructures in various types were extensively studied owing to their superior inherent physico-chemical properties. Whereas excellent electrosorption performance was reported – expressed in terms of salt adsorption capacity (SAC) or average salt adsorption rate (ASAR) – the cost-benefit of graphene-based electrodes, considering total production cost and much lower price of commercial activated carbon, is still controversial. Here, we explore partially exfoliated thermally reduced graphene oxide (GO) – denoted as PE-rGO – prepared by scalable low-temperature thermal exfoliation of GO under air atmosphere. PE-rGO displays a “paper-like” structure with nanoscale pores. By the construction of a lab-scale system, a few grams of product were produced in one batch. A PE-rGO electrode assembled in membrane-CDI three-electrode configuration showed moderate to high SAC of around 13 mg/g under potential window of 0–550 mV versus Ref. electrode in 2000 ppm NaCl solution. However, the energy consumption was shown to be nearly constant with increasing ASAR. This has significant implications for the energy consumption and the projected capital costs.

Original languageEnglish
Article number114599
StatePublished - 15 Oct 2020
Externally publishedYes


  • Capacitive deionization (CDI)
  • Reduced graphene oxide
  • Thermal exfoliation


Dive into the research topics of 'Thermally reduced graphene oxide as an electrode for CDI processes: A compromise between performance and scalability?'. Together they form a unique fingerprint.

Cite this