journal of materiomics - diva portaluu.diva-portal.org/smash/get/diva2:1272288/fulltext01.pdf ·...

ilable at ScienceDirect

Journal of Materiomics xxx (xxxx) xxx

Contents lists ava

Journal of Materiomics

journal homepage: www.journals .e lsevier .com/journal-of -mater iomics/

Polydopamine-based redox-active separators for lithium-ion batteries

Ruijun Pan a, Zhaohui Wang a, *, Rui Sun b, Jonas Lindh b, Kristina Edstr€om a,Maria Strømme b, Leif Nyholm a, **

a Department of Chemistry-Ångstr€om Laboratory, Uppsala University, Box 538, SE-751 21, Uppsala, Swedenb Nanotechnology and Functional Materials, Department of Engineering Sciences, The Ångstr€om Laboratory, Uppsala University, Box 534, SE-751 21,Uppsala, Sweden

a r t i c l e i n f o

Article history:Received 1 October 2018Received in revised form14 November 2018Accepted 14 December 2018Available online xxx

Keywords:Lithium-ion batterySeparatorCellulosePolydopamineRedox-active

* Corresponding author.** Corresponding author.

E-mail addresses: [email protected] (Z. W(L. Nyholm).

Peer review under responsibility of The Chinese C

https://doi.org/10.1016/j.jmat.2018.12.0072352-8478/© 2018 The Chinese Ceramic Society. Pcreativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article as: Pan R et al., Polyddoi.org/10.1016/j.jmat.2018.12.007

a b s t r a c t

The performance of lithium-ion batteries (LIBs) can be effectively enhanced with functionalized sepa-rators. Herein, it is demonstrated that polydopamine-based redox-active (PRA) separators can provideadditional capacity to that of typical anode materials, increase the volumetric capacity of the cell, as wellas, decrease the cell resistance to yield an improved performance at higher cycling rates. The PRA sep-arators, which are composed of a 2 mm thick electrically insulating nanocellulose fiber (NCF) layer and an18 mm thick polydopamine (PDA) and carbon nanotube (CNT) containing redox-active layer, are readilyproduced using a facile paper-making process. The PRA separators are also easily wettable by commonlyemployed electrolytes (e.g. LP40) and exhibit a high dimensional stability. In addition, the pore structureendows the PRA separator with a high ionic conductivity (i.e. 1.06mS cm�1) that increases the rateperformance of the cells. Due to the presence of the redox-active layer, Li4Ti5O12 (LTO) half-cells con-taining PRA separator were found to exhibit significantly higher capacities than the corresponding cellscontaining commercial separators. These results clearly show that the implementation of this type ofredox-active separators constitutes a straightforward and effective way to increase the energy and powerdensities of LIBs.© 2018 The Chinese Ceramic Society. Production and hosting by Elsevier B.V. This is an open access article

under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

The ever-increasing demand for efficient energy storage inportable devices, electrical vehicles and large-scale energy storagesystems has pushed the development of lithium-ion batteries (LIBs)towards the development of devices with higher energy and powerdensities [1e3]. These efforts have so far mainly involved im-provements of the electrode materials via the development of newmaterials and tailoring of their properties [4e6]. Relatively littleattention has, so far, been paid to the development of new orfunctionalized separators, although it is well-known that theproperties of the separators can affect the battery performancesignificantly, mainly due to its influence on the ionic transportpathways and the current density distributions at the electrodes[7,8].

ang), [email protected]

eramic Society.

roduction and hosting by Elsevie

opamine-based redox-active

While contemporary LIB separators are dominantly based onpolyolefin (polypropylene: PP or polyethylene: PE) materials, theirintrinsic drawbacks, e.g. poor wettabilities with respect to con-ventional electrolytes and insufficient thermal stabilities atelevated temperatures [9e12], have led to attempts to developimproved polyolefin separators [13e15] as well as new alternativeseparator materials [16e21]. To enable a utilization of new gener-ations of materials and the realization of energy storage deviceswith high energy and power densities, functionalized separatorshave recently attracted significant attention. Conductive coatings[22,23], ion adsorbing agents [8,24], size excluding layers [25,26],permselective media [27,28], as well as different combinations ofthe latter approaches [29], have hence been used in conjunctionwith commercial separators, or separators based on newmaterials,to alleviate unwanted diffusion of specific ions (e.g. polysulfide ionsand Mn2þ) [30]. Separators composed of several layers have like-wise been developed to tackle safety issues associated with lithiumdendrite formation on lithium metal electrodes via the use offunctionalized intermediate layers enabling early detection [31] orscavenging of dendritic lithium [32]. However, as the majority ofthe aforementioned functionalization approaches resulted in thick

r B.V. This is an open access article under the CC BY-NC-ND license (http://

separators for lithium-ion batteries, Journal of Materiomics, https://

http://creativecommons.org/licenses/by-nc-nd/4.0/

mailto:[email protected]

mailto:[email protected]

www.sciencedirect.com/science/journal/23528478

www.journals.elsevier.com/journal-of-materiomics/

https://doi.org/10.1016/j.jmat.2018.12.007




Fig. 1. Schematic description of the design of the PRA separator. The porous redox-active PDA and CNT containing layer was designed to be in contact with the anode.Due to the capacity contribution from the redox-active (RA) layer, the actual amount ofanode material needed to balance the capacity of the cathode can then be reduced.This allows the volume of the cell to be decreased without affecting the capacity.

R. Pan et al. / Journal of Materiomics xxx (xxxx) xxx2

or densely structured separators that increased the volume of thebattery and/or gave rise to increased cell resistances, the use ofthese separators had an adverse effect on the energy and powerdensities of the batteries. Furthermore, since the latter function-alized separators were generally designed for specific electrodematerials and batteries, general approaches to simultaneously in-crease the energy and power densities of LIBs by merely func-tionalizing the separator still remain scarce.

One possibility would, however, be to use redox active separa-tors, such as the recently described nanocellulose and polypyrrole(PPy) bi-layered separator [33], which can be used with commoncathode materials (i.e. at potentials >2.5 V vs. Liþ/Li). The latterseparator was composed of a 3 mm thick nanocellulose layer,functioning as the true separator layer, and a 7 mm thick porousPPy/nanocellulose composite layer functioning both as a mechan-ical support and a redox-active layer providing additional capacityto the cell. As the total thickness of the redox-active separator canbe made equal to that of the conventional separator, the redoxactive separator can increase the volumetric capacity of the cellsignificantly. Since the aforementioned PPy based redox-activeseparator was designed to operate as an extra cathode material,there is also the corresponding possibility to develop a redox-activeseparator that functions as an additional anode material. This is,unfortunately, complicated by the fact that there are only a ratherlimited number of polymers with redox potentials in the appro-priate potential region. A redox-active separator that can contributeextra capacity in a potential region of common anode materials ishence still to be demonstrated.

During the last decade there has been a growing interest in theuse of renewable and naturally abundant materials in energystorage devices, such as batteries and supercapacitors [34e37]. Thishas generated a large interest in the use of cellulose and its de-rivatives, in e.g. binders [38], gel polymer electrolyte components[39] and flexible substrates [40,41]. Cellulose is presently alsoconsidered a promising alternative materials for LIB separators[42e44], due to its excellent hydrophilicity and thermal stability[10,18,20]. Another interesting renewable material is dopamine,which is a major pigment in natural melanin and which has foundto play an essential role in the adhesion of mussels to solid surfaces[45]. Dopamine has abundant functional groups and can undergooxidative self-polymerization in alkaline aqueous solutions togenerate polydopamine (PDA) with different morphologies (e.g.thin film, spheres) depending on the choice of temperature, oxidantand reaction time [45]. Due to the excellent adhesion ability of PDAto virtually all types of organic and inorganic surfaces, PDAwas firstpresented as a versatile surface coating material in 2007 [46]. Morerecently, PDA has been used in conjunction with energy storagedevices as a carbon source [47] and as a hydrophilic coating on PEseparators [12,13]. Due to its redox-activity, stemming mainly fromthe catechol groups [48], PDA has also been studied both as anodeand cathodematerial [36,49] for LIBs and sodium-ion batteries. ThePDA molecular structure is, unfortunately, complex and still notfully understood due to the complicated polymerization reactionyielding a multitude of different molecules [48]. PDA has, never-theless, been found to exhibit interesting electrochemical behavior[36,49]. Its significant electrochemical activity within the potentialregion of interest in conjunction with anode materials and excel-lent adhesion tomost surfaces make PDA a promising candidate foruse in a redox-active cellulose based separator compatible with LIBanodes.

Herein, a new bilayer-structured redox-active separator,comprising a thin NCF layer and a thicker NCF/PDA/CNT compositelayer, is manufactured and used to increase the capacity of the LIBanode. The polydopamine based redox-active (PRA) separatorconsists of a 2 mm thick NCF separator layer and an 18 mm thick

Please cite this article as: Pan R et al., Polydopamine-based redox-activedoi.org/10.1016/j.jmat.2018.12.007

redox-active layer containing the PDA particles, CNTs and NCFs,where both layers are hydrophilic and thermally stable. The 2 mmthick electrically insulating NCFs layer, which thus serves as thetrue separator layer, merely represents 8% of the volume occupiedby a conventional 25 mm thick polyolefin separator. Meanwhile, the18 mm thick highly porous redox-active layer acts both as a sup-porting layer, providing the required mechanical strength andflexibility, as well as a source of additional capacity in a potentialwindow compatible with common LIB anode materials (i.e. 0e3 Vvs. Liþ/Li).

The performances of cells containing LTO and lithium metal (Li)electrodes in addition to either a PRA or a conventional PE separatordemonstrate that the PRA cell exhibited higher capacities than thecorresponding PE cell at all studied cycling rates. This can beexplained by the higher ionic conductivity of the PRA separator andthe redox-active properties of the PDA containing layer acting as anextension of the LTO anode. The main components of the PRAseparator are biosourced and biodegradable, as the NCFs originallywere extracted directly from sea algae by our group [50,51], whiledopamine is a naturally occurring biomolecule employed in manydifferent fields [46,52e54]. As the production of the PRA separatoronly involves self-polymerization of dopamine, sonication of NCFsand CNTs, as well as low vacuum filtration, the PRA separator ismore environmentally friendly than the conventional fossil fuelderived polyolefin separators, particularly as the low-energy-demanding process readily can be scaled up. The present PRAseparator approach therefore constitutes a new promising routetowards the realization of more sustainable LIBs with increasedenergy and power densities.

2. Results and discussion

Composition, morphology and conductivity. As indicated inFig. 1, the PRA separat or has a bi-layered structure with a pure NCFlayer and a redox-active PDA/NCF composite layer. Scanning elec-tron microscopy (SEM) images of the PRA separator, as schemati-cally indicated in Fig. 2, confirmed the successful realization of thebi-layered structure as two distinct layers can be observed in Fig. 3awith the thin NCF layer and the porous redox-active PDA/NCF layerabove and below the red dashed line, respectively. X-ray photo-electron spectroscopy (XPS) results (see Fig. 3e) clearly indicatedthe presence of N-containing species in the redox-active layer,which can be attributed to the presence of the PDA components.The presence of the PDA components was also verified in an Energy


Fig. 2. Schematic illustration of the process used in the manufacturing of the PRA separator. The PDA/NCFs dispersion was prepared by self-polymerization of dopamine monomersin an alkaline solution also containing NCFs and oxygen. A SEM image of the obtained redox-active (RA) PDA/CNT/NCF composite is shown in Fig. S5.

R. Pan et al. / Journal of Materiomics xxx (xxxx) xxx 3

Dispersive X-ray spectroscopy (EDX) element mapping analysis ofthe redox-active layer (see Fig. S1), demonstrating that the nitrogen(as well as the carbon and oxygen) was distributed homogenouslywithin the studied region.

As discussed in the literature [48], the PDA generating self-polymerization process is complicated by multi-step reactionsand still not fully understood. It has, however, been proposed [48]that the PDA molecules mainly consist of covalently linked 5,6-dihydroxyindole (DHI) and 5,6-indolequinone (IQ) units (seeFig. 2), since a fraction of the DHI molecules can be oxidized to IQmolecules during the process [36,49]. Characteristic signals fromthe C¼O groups in the IQ units were indeed observed in the XPS C1s and O 1s spectra, as well as in the Fourier transform infraredspectroscopy (FTIR) spectra for the redox-active layer (see Fig. S2).

SEM images of the PRA separator also show that the thin butrelatively compact NCF layer had the previously reported structure(see Fig. 3b) [20,33,55], whereas the redox-active layer featured amore porous structure with intertwined NCFs and CNTs forming a3D network inwhich spherical PDA particles were present as fillers.SEM images of a NCF separator, a CNT/NCF sample and a PE sepa-rator are shown in Fig. S3. In the PRA design, the thin NCFs layerthus acts as the insulating layer that physically prevents directcontact of the anode and cathode in the battery, whereas the moreporous redox-active layer provides the necessary mechanical sup-port to the thin NCFs layer and also contributes to the energystorage capacity.

To demonstrate the mechanical functionality of the redox-activelayer, a 3 mm thick NCF film (see Fig. S4) was also produced with thesame paper-making method. This thin and unsupported NCF filmwas, however, difficult to peel off the filter membrane due to itsstrong interaction with the latter and its insufficient mechanicalstrength. Even though a 5 mm thick NCFs film could be produced,the handling of such a thin NCF film was rather difficult due toelectrostatic forces and its lowmechanical strength. These findings,which clearly show the problems associated with the use of NCF-


based separators with a thickness of up to 5 mm, indicate that it ispractically difficult to increase the volumetric capacity of the bat-teries by employing such a thin separator. This problem is cir-cumvented using the PRA approach as the redox-active PDAcontaining layer also serves as a mechanical support for the NCFseparator layer that can then be made as thin as 2 mm. Thisconsequently means that the electro-inactive volume of the PRAseparator can be reduced to only that of the 2 mm thick NCFs layer. Ifthe capacity of redox-active layer can indeed be employed, theenergy density can consequently be increased compared to that fora conventional cell comprising a ~25 mm thick conventionalseparator.

It is worth noting that the porous redox-active layer not onlyshould provide efficient electron transfer pathways due to theincluded CNTs but also should facilitate the Liþ mass transportrequired by the PDA electrochemical reactions. Since the PDAcontaining redox-active layer practically serves as an additionalanode material present it could be argued that the same effectcould be obtained by using a thinner conventional separator and athicker layer of the anode material. As already indicated, the use of5 mm thin separators is, however, not straightforward. Moreover,the use of a thicker anode material would not necessarily result inthe expected increase in the capacity of the anode due to theincreased diffusion length in the anode material. As the diffusiontime is proportional to the square of the diffusion length (e.g. thethickness of the anode material), an increase of the electrodethickness by a factor of e.g. two would thus lead to a four-fold longdiffusion time (e.g. cycling time) in order to access the total ca-pacity. However, as the PRA separator is porous, the mass transportwithin the PDA containing layer is not an issue, which means thatits inherent capacity should be readily accessed even at high cyclingrates. Note also that the electrochemical reactions involving theconventional anode material should remain unaffected by the PDAcontaining layer, at least not in an electrolyte containing e.g. 1 MLiPF6, due to the large excess of lithium ions.


Fig. 3. Morphology and composition of the PRA separator. (a) Cross-sectional SEM image of the PRA separator in which the red dashed line roughly mark the boundary between theNCFs layer (above the line) and the redox-active PDA and CNT containing layer (below the line). The inset shows a digital image of the PRA separator. (b) Surface SEM image of theNCFs layer. (c and d) Surface SEM images of the redox-active layer at different magnifications. (e) N1s XPS spectra for the redox-active layer of the PRA separator and the CNT/NCFcontrol sample, respectively. The inset shows the corresponding survey XPS spectra.


The highly porous redox-active layer of the PRA separatorshould likewise be beneficial to decrease the cell resistance. Thepore structure of a 20 mm thick PRA separator, which is described inFig. 4, featured well-distributed large pores with an average size of35 nm while the corresponding value for a 20 mm thick NCFmembrane was 13 nm. More importantly, when compared with aconventional PE separator, Solupor®, the PRA separator exhibited asignificantly higher specific pore volume in the studied region (seeFig. 4a, the areas under the curves), a higher porosity (32% vs. 20%)as well as a higher ionic conductivity (1.06 vs. 0.17mS cm�1,determined using ac impedance experiments after soaking withLP40 electrolyte), as can be seen in Table S1. The larger porosity,yielding a higher ionic conductivity and a lower cell resistance forthe same separator thickness, should clearly be of significant


importance to the rate performance of the LIBs. It can hence beconcluded the LIB performance is expected to be enhanced withrespect to the energy and power densities when replacing a con-ventional polyolefin separator with the proposed PRA separator.This conclusion is also supported by the electrochemical resultsdescribed below.

Thermal stability and electrolyte wettability. Although thepoor thermal stabilities and wettabilities of commercial polyolefinseparators generally are tolerated in contemporary LIBs, mucheffort has been made to address these issues with polyolefin sep-arators, as well as to search for better alternative materials [7,13].The thermal stability is naturally very important to the batterysafety at elevated temperatures, which makes this a particularlyimportant question for batteries designed for high power


Fig. 4. Pore size distributions for the PRA, NCF and PE separators as well as Nyquist plots for symmetric cells (Cu/separator/Cu) cells equipped with PRA and PE separators,respectively. The high frequency intercepts in the Nyquist plots (see the insets) were used to calculate the ionic conductivities. The semicircle seen for the PRA separator stemmedfrom the PDA and CNT containing redox-active layer.


applications. The wettability, which is coupled to the easiness ofelectrolyte incorporation into the pores of the separator, is of sig-nificant practical importance in the battery manufacturing processand naturally also depends on the choice of electrolyte [10]. Fig. 5shows the results of thermal stability and wettability tests carriedout with the PRA and PE separators. In the thermal stability test, inwhich the temperature was increased from room temperature to150 �C, it was found that the size of the PE separator decreasedsignificantly during the test while the PRA separator maintaineddimensionally intact (Video S1). In the wettability test, where LP40electrolyte was added drop by drop with a syringe onto the sepa-rators, it was found that the drops were absorbed more quickly bythe PRA separator. This indicates that the PRA separator had ahigher affinity for the LP40 electrolyte than the PE separator, as canbe more clearly seen in the videos recorded during the experiment(Video S2). The good thermal stability and high wettability hencemake the PRA separator a promising alternative to the commercialpolyolefin LIB separators.

Supplementary video related to this article can be found athttps://doi.org/10.1016/j.jmat.2018.12.007.

Electrochemical performance. Asmentioned above, the redox-active PDA containing layer should contribute to the charge storagecapacity within a potential region compatible with those of typical

Fig. 5. Digital images takes before and after the thermal stability (left) and wettability testsfurther studied in the videos found in the supporting information (i.e. Video S1 and S2).


LIB anode materials. To validate the PRA separator design and toevaluate the charge storage capacity of the PDA based redox-activelayer, electrochemical measurements (i.e. cyclic voltammetric (CV)and galvanostatic (GS) experiments) were carried out with the PRAseparator as one of the electrodes. In these experiments the CNT toPDA mass ratio in the redox-active layer was ~2:5. The employedtwo-electrode cells comprised a Li electrode and an electrolytesoaked PRA separator with the NCF layer in direct contact with theLi electrode. This cell configuration is referred to as the Li/PRA cellbelow.

Fig. 6a shows cyclic voltammograms depicting the first threecycles for the Li/PRA cell, recorded in a potential window between0.1 and 3 V vs. Liþ/Li at a scan rate of 0.1mV s�1. In agreement withthe results of Sun et al. [36], a large irreversible discharge capacitywas seen on the first cycle (see the reduction peak at about 0.7 vs.Liþ/Li) while relatively stable cycling was observed on the subse-quent cycles. The initial irreversible capacity can be attributed to areduction of surface groups and the formation of a solid electrolyteinterphase (SEI) layer on the CNTs [38]. This is further supported bythe large irreversible reduction capacity seen during the first fivecycles for a CNT/NCF sample in Fig. S6. On the first anodic scan, twosmall and broad bumps can, nevertheless, be seen at about 1.7 and2.5 V vs. Liþ/Li, most likely due to surface confined oxidation

(right) for the PRA (black) and PE (white) separators. The wettability tests can also be



Fig. 6. Electrochemical performance of the PRA separator in the potential window between 0.1 and 3 V vs. Liþ/Li. (a) Cyclic voltammograms showing the first three cycles for theemployed Li/PRA cell using a scan rate of 0.1mV s�1. (b) The specific charge/discharge capacities of the Li/PRA cell as a function of the cycle number during 280 cycles obtained witha current density was 180mA g�1

PDA. The voltage versus time profiles in the inset show the discharge and charge curves for the 50th cycle. (c) Voltage versus specific capacity curvesfor the Li/PRA cell depicting the 1st, 10th, 50th, 100th and 200th cycles. (d) Specific charge and discharge capacities of the Li/PRA cell as a function of the cycle number and currentdensity (i.e. 90, 180, 360, 720mA g�1

PDA).


reactions. The increasing cathodic current at potentials belowabout0.5 V vs. Liþ/Li can most likely be ascribed to the lithiation of theCNTs. The absence of a corresponding Li oxidation peak suggeststhat the corresponding delithiation step was incomplete, in goodagreement with previous results for other carbon based electrodes[56e58]. As this phenomenon, which often is referred to as lithiumtrapping, was also seen on the second and third cycles, it appears asif the surfaces of the CNTs were partially re-oxidized on the anodicscan [56e58]. A small oxidation peak can also be seen at about0.25 V vs. Liþ/Li predominately on the second and third scans. Theshapes of the second and third voltammetric cycles indicate thatthe PRA separator generally exhibited a capacitive behavior anal-ogous to that seen for many carbonaceous materials.

Based on the second voltammetric cycle, the integrateddischarge (i.e. cathodic) and charge (i.e. anodic) capacities of thePRA separator were approximately 375 and 309 mAh g�1

PDA (basedon the PDA mass), excluding the capacity contribution from theCNTs (which includes double layer charging as well as faradaicreactions). The corresponding capacities, including the CNT ca-pacity contributions were 452 and 375 mAh g�1

PDA. This conse-quently indicates that the capacity contribution from the PDA wasabout 80% both during charge and discharge in these voltammetricexperiments.

The results of the galvanostatic cycling tests (see Fig. 6b) were


generally consistent with the voltammetric results, as the firstdischarge of the PRA separator was associated with a large irre-versible capacity. However, it was noted that the overall capacity(due to both the PDA and CNTs) calculated from the galvanostaticresults (~200 mAh g�1

PDA) was much lower here than that in thevoltammetric experiments (~452 mAh g�1

PDA). This effect was,most likely, due to the different experimental time domains as thetime for a full voltammetric scan was ~16 h, whereas the corre-sponding time in the galvanostatic case was only ~2.3 h (see Fig. 6binset). This would imply that the extent of the redox reactions ofthe redox-active layer were time-dependent.

When comparing Fig. 6b and S6, it can also be seen that thespecific capacities of the PRA separator and the CNT/NCF layer wereapproximately the same (i.e. about 150 mAh g�1 for a currentdensity of 180mA g�1) in the galvanostatic experiments. Thistogether with the fact that the CNT to PDA mass ratio in the redox-active layer was ~2:5 indicates that about 60% of the capacity of thePRA separator stemmed from the PDA. It is hence clear that thecapacity contribution from the PDAwas smaller in the galvanostaticthan in the potentiostatic experiments, suggesting that the capacityof the PDA was more sensitive to the experimental time domainthan the capacity due to the CNTs. The latter is not unexpected asthe capacity of the CNTs stems from a combination of double layercharging and surface confined redox reactions, which both should



be compatible with the use of high scan rates. Since the PDA waspresent as particles with a size of about 100 nm, the observedinability to access the same capacity at increased cycling rates issomewhat puzzling. It, however, appears as if a passivating layer isformed on the PDA particles, predominantly during the chargingsteps (i.e. during the oxidation of the PDA). Interestingly, the gal-vanostatic results also indicated the presence of an initial capacityincrease for the Li/PRA cell, which could be explained by an acti-vation of the PDA spheres via an increase of their accessible surfacearea. The fact that this increase was predominantly seen for thecharge capacity could be explained by the additional presence ofthe irreversible reactions taking place during the initial dischargesteps.

As seen in Fig. 6b and c, there was a gradual decrease in the PRAcapacity during prolonged cycling whichmost likely stemmed froma degradation of the electrochemical performance of the PDA.Nevertheless, the results still demonstrate that the PRA separatorexhibited a capacity of about 180 mAh g�1

PDA in the potentialwindowbetween 0.1 and 3 V vs. Liþ/Li at a cycling current density of180mA g�1

PDA. This specific capacity, which incidentally isapproximately half of that of graphite [59,60], shows that PDA iswell-suited for use in this type of redox-active separators. More-over, the PRA separator exhibited an even higher capacity at lowerrates (~250mAh g�1

PDA at 90mA g�1PDA) andmaintained a capacity

of ~100 mAh g�1PDA at a current density of 720mA g�1

PDA (seeFig. 6d). It can hence be concluded that the PRA separator can giverise to a significant capacity addition in the voltage region of in-terest when using common LIB anode materials. This is furtherdemonstrated in the next section.

Electrochemical performance of a proof-of-concept cell. Thebasic idea when using the PRA separator is thus that the redox-active layer acts as both a part of the separator and as an exten-sion of the anode. In this way it is possible to minimize the electro-inactive volume within the cell and to increase its volumetric (andgravimetric) energy and power densities. To validate the conceptthe electrochemical performances of LTO/Li cells containing eithera PE or a PRA separator were compared as is shown in Fig. 7. Itshould be noted that in the LTO/PRA/Li cell the redox-active layerwas in direct contact with the LTO electrode. In Fig. 7a it can be seenthat the PRA cell constantly exhibited higher capacities than the PEcell (i.e.170 vs.150mAh g�1 based on the LTOmass) at a cycling rateof 1 C. Comparisons of capacities based on the LTO weight wereused for simplicity as both cells were equipped with identical LTO(and Li) electrodes. In analogy with the Li/PRA results discussedabove, a large irreversible capacity was seen for the PRA cell but inthis case the charge and discharge capacities increased rather thandecreased somewhat during the cycling. The presence of initialirreversible reactions is also reflected in the Coulombic efficiency(CE) values for the PRA cell which increased during the cycling andstabilized at the same level as for the PE cell after about 20 cycles.

When scrutinizing the voltage versus specific capacity curves inFig. 7b, it can further be seen that the PRA cell exhibited a largercapacity than the PE cell in the 1.5e3 V potential window. The lattercan clearly be attributed to the presence of the PDA and CNT con-taining redox-active layer, and higher capacities were also seen forthe PRA cell when increasing the cycling rate (see Fig. 7c). Thelargest difference between the PRA and PE cells was, however, seenat the lowest cycling rate in good agreement with the decreasingPDA capacity at increasing cycling rates discussed above. It is alsoclear that the PRA cell exhibited lower overpotentials both duringcharge and discharge which resulted in more well-defined plateausthan for the PE cell. These differences between the shapes of thevoltage versus specific capacity curves, which became even morepronounced at higher cycling rates (see Fig. 7 d and e), can beascribed to the lower electrolyte resistance of the PRA separator


(see the high frequency intercepts in Fig. 6f). Moreover, the PRA cellexhibited a capacity of ~210 mAh g�1

LTO at 0.5 C (vs. ~150 mAh g�1

LTO for the PE cell) and ~110 mAh g�1LTO at 5 C (vs. ~70 mAh g�1

LTOfor the PE cell), confirming the enhanced rate capability of the cellwith PRA separator seen in Fig. 7c. It should be noted that thetheoretical capacity of LTO is 170 mAh g�1

LTO and that the highervalues presented above (e.g. 210 mAh g�1

LTO) hence must beascribed to the capacity contribution from the redox-active layer inthe PRA separator.

The capacity contributions to the PRA separator from the PDAand CNTs were also evaluated at different cycling rates. In theemployed potential window (i.e. 1e3 V, see Fig. S8), the CNTsfeatured a significantly lower capacity than in the 0.1e3 V region(i.e. ~30 versus ~185 mAh g�1

CNT at 90mA g�1PDA). The additional

capacity due to the PRA separator in the LTO/PRA/Li proof-of-concept cell must consequently mainly be ascribed to the PDAsince PDA was the major redox-active component in the redox-active layer (since mPDA:mCNT¼ 5:2) and the CNTs only had a ca-pacity of ~30 mAh g�1

CNT at 90mA g�1PDA. In fact, the lower CNT

capacity in the voltage range of 1e3 V is not surprising, given thesignificant capacity contribution seen at voltages below about 0.5 Vin Fig. 6a and c. Regarding the capacity contribution from the PRAseparator, it is important to mention that the magnitude of theabsolute capacity increase depends strongly on the amounts of theanode material and the PDA as well as the employed voltage win-dow. This means that the effects on the energy and power densitiesof different LIBs upon the replacement of the conventional sepa-rator with a PRA separator need to be determined for each indi-vidual LIB. The present results, nevertheless, clearly demonstratethe feasibility of the present redox-active separator concept, itspotential usefulness when it comes to decreasing the electro-inactive volume within the LIBs, and the possibility to enhancethe energy and power densities of LIBs by merely exchanging theseparator.

3. Conclusions

The development of future high-performance LIBs will mostlikely depend significantly on the development of new advancedseparators since the use of functionalized separators constitutes astraightforward way of improving the battery performance. In thepresent work, a bi-layered polydopamine based redox-activeseparator (PRA), compatible with the potential region of typicalanode materials, was designed and manufactured using biosourcedand sustainable materials (i.e. nanocellulose and polydopamine)and a facile water-based paper-making process. By minimizing thethickness of the nanocellulose-based insulating layer to 2 mm andadding an 18 mm thick polydopamine and carbon nanotube con-taining redox-active layer as mechanical support, a new function-alized separator could be manufactured. Due to the thin insulatinglayer and the additional capacity provided by the redox-activelayer, the capacity of a LIB could be increased merely by replacingthe conventional separator without affecting the thickness of thecell or the electrode reactions at the anode and cathode.

The highly porous redox-active layer, which is electroactivebetween 0.1 and 3 V vs. Liþ/Li, thus serves as an extension of theanode and can therefore be used irrespective of the choice of theanode material. While morphological and compositional evalua-tions demonstrated a successful manufacture of the PRA separator,further assessments also confirmed advantages of the PRA sepa-rator with respect to a commercial PE based separator regardingthe thermal stability, wettability and ionic conductivity. Resultsobtained with a proof-of-concept LTO/Li cell further showed thatthe use of the PRA separator resulted in an enhanced energy stor-age capacity and rate capability compared to the corresponding cell


Fig. 7. (a) The specific charge and discharge capacities and coulombic efficiency as a function of the cycle number for the LTO/Li cells equipped with either a PRA or PE separator forcycling at a rate of 1 C. (b) Voltage versus specific capacity curves depicting the 10th cycles for the PRA and PE separator containing LTO/Li cells at a rate of 1 C. (c) Specific charge/discharge capacities of LTO/Li cells with PRA and PE separators, respectively, as a function of the cycle number and cycling rate (i.e. 0.5 C, 1 C, 2 C, 5 C). (d and e) Voltage versusspecific capacity curves obtained with the indicated cycling rates using LTO/Li cells containing PRA and PE separators, respectively. (f) Nyquist plots for LTO/Li cells with PRA and PEseparators, respectively, where the inset shows the high frequency regions of the Nyquist plots. The cycling rates were calculated based on the capacity of the LTO electrode.


containing a conventional PE separator. Bymaking themajor part ofthe LIB separator redox-active, the PRA approach not only providesa promising means for the simultaneous attainment of enhancedenergy and power densities of LIBs, it also paves ways for the designof functionalized separators containing other interesting materials.

4. Experimental section

4.1. Materials

Cladophora cellulose powder (Batch NO. G3095-10), producedby the Cladophora green algae, was supplied by FMC BiopolymerUSA. Dopamine hydrochloride, ammonia aqueous solution, multi-walled carbon nanotubes, lithium titanium oxide, Ketjen black,1.0M lithium hexafluorophosphate (LiPF6) in ethylene carbonate(EC)/diethyl carbonate (DEC) (1/1, v/v, BASF, i.e. LP40), Li foil(125 mm, Cyprus Foote Minerals) Solupor® (E�8P01E, DSM), nylonfilter membrane (F90mm, 0.45 mm, Magna) were purchased fromtheir indicated sources and used as received. The other materials(e.g. deionized water, ethanol, aluminum foil, copper foil) were allbought from commercial manufacturers and used without anyfurther purification.

4.2. Preparation of PDA/NCF dispersion

In the preparation of the PDA/NCF dispersion, 200mg Clado-phora cellulose powder was ultra-sonicated with 70mL deionizedwater in a sonication beaker for 15min (VibraCell 750W, Sonics,USA). The homogenous NCFs dispersion was then transferred to aflask where 40mL ethanol, 20mL deionized water and 2mLammonia aqueous solution had been pre-mixed for 20min. Thedispersion in the flask was stirred to allow homogenization foranother 30min prior to the drop by drop addition of a dopamine


hydrochloride (DH) solution, prepared by dissolving 500mg DH in10mL deionized water. The color of the mixture was immediatelychanged to salmon and then gradually to dark brown. The reactionwas allowed to proceed for 24 h before using the PDA/NCF disper-sion in the preparation of the redox-active layer. Another samplewithout any NCFs (but with the same amounts of solvents andother reagents) was also prepared for comparison. The latter wasdone to illustrate that when the NCFs were used in the synthesis,PDA spheres with smaller sizes could be obtained (see Fig. S5). Acontrol sample with 10mg CNTs and 50mg NCFs was likewiseproduced with similar method (denoted as CNT/CNF).

Preparation of the PRA separator. The production of the PRAseparator is schematically illustrated in Fig. 2. A 2mgmL�1 NCFsdispersion in deionized water was prepared by ultra-sonication.4mL of this dispersion was then diluted to 40mL with deionizedwater. The PDA/CNT/NCF dispersion was then prepared by mixing10mL of the aforementioned PDA/NCF dispersion, 8mg CNTs, 4mgNCFs in 20mL ethanol and 30mL deionized water, followed byultra-sonication for 15min. The PRA separator was then manufac-tured by filtering the PDA/CNTs/NCFs dispersion on top of a filteredNCFs layer at a reduced pressure. The obtained wet PRA membranewas dried at room temperature overnight in a Petri dish, followedby heat treatment at 150 �C for 12 h. The PRA membrane wassubsequently cut into ø20mm discs and dried in a vacuum furnaceat 120 �C overnight before further use. The PDAmass fraction in thePRA separator was determined by weighing the latter, based on theknown masses of the other components.

Characterization of the PRA separator. The morphology of thesamples was studied with a field emission scanning electron mi-croscope (Zeiss LEO1550, Germany) and the element mapping ofthe redox-active layer was obtained using EDX. The XPS spectrawere obtained with a Physical Electronics Quantum 2000 ScanningESCA spectrometer equipped with monochromatized Al Ka



radiation (hn¼ 1486.7 eV). FTIR characterizations were performedwith a Bruker FTS66v/s spectrometer using an attenuated totalreflectance sample holder at a resolution of 4 cm�1. The nitrogensorption analyses were performed at �194 �C to examine themesopores (i.e. pores with diameter of 2e50 nm [61]) of the sep-arators with a surface area analyzer (Micromeritics ASAP2020). Thepore size distributions were determined with the Barrett-Joyner-Halenda (BJH) method based on the desorption processes. Theporosities of the separators were calculated based on theirapparent and bulk densities.

4.3. Electrode and cell preparation

LTO electrodes were prepared by first mixing LTO, carbon blackand polyvinylidene fluoride powders in a weight ratio of 80:10:10with N-Methyl-2-pyrrolidone as the solvent in a ball-milling jar.The slurry was then intensely mixed by planetary ball-milling for1.5 h and coated onto an aluminum foil. The electrode foil was thenallowed to dry at 100 �C in the coating machine for 15min before itwas cut into ø13mm electrode discs and transferred into an argon-filled glove box and further dried at 120 �C overnight. Lithiummetalwas cut into ø14mm circular discs which then were used as com-bined counter and reference electrodes. The bare copper andaluminum electrodes (ø13mm discs) were prepared and cleanedwith ethanol before transfer into the glovebox. Two-electrodepouch cells were assembled by sandwiching the electrolyte-soaked separator between the LTO electrode and the Li foil priorto sealing with a sealing machine in the argon-filled glovebox(water content <1 ppm, oxygen content <1 ppm).

Electrochemical analyses. Electrochemical impedance spec-troscopy (EIS) measurements in the frequency range between 10�1

and 105 Hz were conducted to measure the impedance of sym-metric (copper/separator/copper) cells at the open circuit voltageusing a VMP instrument (Biologic Multichannel potentiostat). Theapplied amplitude of the ac voltage was 10mV. The high frequencyintercepts of the Nyquist plots were used to calculate the ionicconductivities of the separators. EIS measurements were also car-ried out for the LTO/Li cells with different separators following thesame testing protocols. The cyclic voltammetric measurementswere performed on the same VMP instrument using a potentialscan rate of 0.1mV s�1 for all cells. The galvanostatic cycling,including the rate tests, on the Li/PRA and LTO/Li cells with differentseparators were performed on an Arbin instrument (model BT-2043).

Author contributions

The manuscript is written through contributions of all authors.All authors have given approval to the final version of themanuscript.

Notes

The authors declare no competing final interest.

Acknowledgments

The authors would like to acknowledge X. Xu and C. Xu for theassistance with XPS data collection and interpretation, C. Ruan forthe help with FTIR measurement and J. Zhao for drawing self-polymerization scheme of dopamine. The work is financially sup-ported by The Swedish Energy Agency (i.e. the TriLi project andproject 2017-013543) as well as StandUp for Energy.


Appendix A. Supplementary data

Supplementary data to this article can be found online athttps://doi.org/10.1016/j.jmat.2018.12.007.

References

[1] Tarascon JM, Armand M. Issues and challenges facing rechargeable lithiumbatteries. Nature 2001;414:359e67.

[2] Etacheri V, Marom R, Elazari R, Salitra G, Aurbach D. Challenges in thedevelopment of advanced Li-ion batteries: a review. Energy Environ Sci2011;4:3243.

[3] Kang K, Meng YS, Br�eger J, Grey CP, Ceder G. Electrodes with high power andhigh capacity for rechargeable lithium batteries. Science 2006;311:977e80.

[4] Li H, Wang Z, Chen L, Huang X. Research on advanced materials for Li-ionbatteries. Adv Mater 2009;21:4593e607.

[5] Bruce PG, Scrosati B, Tarascon J-M. Nanomaterials for rechargeable lithiumbatteries. Angew Chem Int Ed 2008;47:2930e46.

[6] Scrosati B. Recent advances in lithium ion battery materials. Electrochim Acta2000;45:2461e6.

[7] Arora P, Zhang ZJ. Battery separators. Chem Rev 2004;104:4419e62.[8] Oh Y-S, Jung GY, Kim J-H, Kim J-H, Kim SH, Kwak SK, Lee S-Y. Janus-faced,

dual-conductive/chemically active battery separator membranes. Adv FunctMater 2016;26:7074e83.

[9] Lee H, Yanilmaz M, Toprakci O, Fu K, Zhang X. A review of recent de-velopments in membrane separators for rechargeable lithium-ion batteries.Energy Environ Sci 2014;7:3857e86.

[10] Zhang SS. A review on the separators of liquid electrolyte Li-ion batteries.J Power Sources 2007;164:351e64.

[11] Huang XS. Separator technologies for lithium-ion batteries. J Solid StateElectrochem 2011;15:649e62.

[12] Ryou MH, Lee YM, Park JK, Choi JW. Mussel-inspired polydopamine-treatedpolyethylene separators for high-power Li-ion batteries. Adv Mater2011;23:3066e70.

[13] Dai J, Shi C, Li C, Shen X, Peng L, Wu D, Sun D, Zhang P, Zhao J. A rationaldesign of separator with substantially enhanced thermal features for lithium-ion batteries by the polydopamine-ceramic composite modification of poly-olefin membranes. Energy Environ Sci 2016;9:3252e61.

[14] Wang Z, Guo F, Chen C, Shi L, Yuan S, Sun L, Zhu J. Self-assembly of PEI/SiO2 onpolyethylene separators for Li-ion batteries with enhanced rate capability.ACS Appl Mater Interfaces 2015;7:3314e22.

[15] Fu D, Luan B, Argue S, Bureau MN, Davidson IJ. Nano SiO2 particle formationand deposition on polypropylene separators for lithium-ion batteries. J PowerSources 2012;206:325e33.

[16] Wei X, Nie Z, Luo Q, Li B, Chen B, Simmons K, Sprenkle V, Wang W. Nano-porous polytetrafluoroethylene/silica composite separator as a high-performance all-vanadium redox flow battery membrane. Adv. EnergyMater. 2013;3:1215e20.

[17] Janakiraman S, Surendran A, Ghosh S, Anandhan S, Venimadhav A. Electro-active poly(vinylidene fluoride) fluoride separator for sodium ion battery withhigh coulombic efficiency. Solid State Ionics 2016;292:130e5.

[18] Chun S-J, Choi E-S, Lee E-H, Kim JH, Lee S-Y, Lee S-Y. Eco-friendly cellulosenanofiber paper-derived separator membranes featuring tunable nanoporousnetwork channels for lithium-ion batteries. J Mater Chem 2012;22:16618.

[19] Zhang B, Wang Q, Zhang J, Ding G, Xu G, Liu Z, Cui G. A superior thermostableand nonflammable composite membrane towards high power battery sepa-rator. Nano Energy 2014;10:277e87.

[20] Pan RJ, Cheung O, Wang ZH, Tammela P, Huo JX, Lindh J, Edstrom K,Stromme M, Nyholm L. Mesoporous Cladophora cellulose separators forlithium-ion batteries. J Power Sources 2016;321:185e92.

[21] Wang Z, Pan R, Xu C, Ruan C, Edstr€om K, Strømme M, Nyholm L. Conductingpolymer paper-derived separators for lithium metal batteries. Energy StorageMater 2018;13:283e92.

[22] Su YS, Manthiram A. Lithium-sulphur batteries with a microporous carbonpaper as a bifunctional interlayer. Nat Commun 2012;3:1166.

[23] Chung S-H, Manthiram A. Bifunctional separator with a light-weight carbon-coating for dynamically and statically stable lithium-sulfur batteries. AdvFunct Mater 2014;24:5299e306.

[24] Kim JH, Gu M, Lee do H, Kim JH, Oh YS, Min SH, Kim BS, Lee SY. Functionalizednanocellulose-integrated heterolayered nanomats toward smart batteryseparators. Nano Lett 2016;16:5533e41.

[25] Bai S, Liu X, Zhu K, Wu S, Zhou H. Metaleorganic framework-based separatorfor lithiumesulfur batteries. Nat. Energy 2016;1:16094.

[26] Bai S, Zhu K, Wu S, Wang Y, Yi J, Ishida M, Zhou H. A long-life lithium-sulphurbattery by integrating zinc-organic framework based separator. J Mater ChemA 2016;4:16812e7.

[27] Huang J-Q, Zhuang T-Z, Zhang Q, Peng H-J, Chen C-M, Wei F. Permselectivegraphene oxide membrane for highly stable and anti-self-dischargelithiumesulfur batteries. ACS Nano 2015;9:3002e11.

[28] Zeng F, Jin Z, Yuan K, Liu S, Cheng X, Wang A, Wang W, Yang Y-s. High per-formance lithium-sulfur batteries with a permselective sulfonated acetyleneblack modified separator. J Mater Chem A 2016;4:12319e27.

[29] Xu Q, Hu GC, Bi HL, Xiang HF. A trilayer carbon nanotube/Al2O3/



http://refhub.elsevier.com/S2352-8478(18)30129-1/sref1











































































































polypropylene separator for lithium-sulfur batteries. Ionics 2015;21:981e6.[30] Xiang Y, Li J, Lei J, Liu D, Xie Z, Qu D, Li K, Deng T, Tang H. Advanced separators

for lithium-ion and lithium-sulfur batteries: a review of recent progress.ChemSusChem 2016;9:3023e39.

[31] Wu H, Zhuo D, Kong D, Cui Y. Improving battery safety by early detection ofinternal shorting with a bifunctional separator. Nat Commun 2014;5:5193.

[32] Liu K, Zhuo D, Lee HW, Liu W, Lin D, Lu Y, Cui Y. Extending the life of lithium-based rechargeable batteries by reaction of lithium dendrites with a novelsilica nanoparticle sandwiched separator. Adv Mater 2016;29:1603897.

[33] Wang Z, Pan R, Ruan C, Edstr€om K, Strømme M, Nyholm L. Redox-activeseparators for lithium-ion batteries. Adv Sci 2018;5:1700663.

[34] Zhu H, Luo W, Ciesielski PN, Fang Z, Zhu JY, Henriksson G, Himmel ME, Hu L.Wood-derived materials for green Electronics, biological devices, and energyapplications. Chem Rev 2016;116:9305e74.

[35] Wang Z, Tammela P, Strømme M, Nyholm L. Cellulose-based supercapacitors:material and performance considerations. Adv Energy Mater 2017;7:1700130.

[36] Sun T, Li Zj, Wang Hg, Bao D, Meng Fl, Zhang Xb. A biodegradable polydop-amine-derived electrode material for high-capacity and long-life lithium-ionand sodium-ion batteries. Angew Chem 2016;128:10820e4.

[37] Renault S, Oltean VA, Araujo CM, Grigoriev A, Edstr€om K, Brandell D. Super-lithiation of organic electrode materials: the case of dilithium benzenedi-propiolate. Chem Mater 2016;28:1920e6.

[38] Wang Z, Xu C, Tammela P, Huo J, Strømme M, Edstr€om K, Gustafsson T,Nyholm L. Flexible freestanding Cladophora nanocellulose paper based Sianodes for lithium-ion batteries. J Mater Chem A 2015;3:14109e15.

[39] Lee JM, Nguyen DQ, Lee SB, Kim H, Ahn BS, Lee H, Kim HS. Cellulose triacetate-based polymer gel electrolytes. J Appl Polym Sci 2010;115:32e6.

[40] Nystr€om G, Razaq A, Strømme M, Nyholm L, Mihranyan A. Ultrafast all-polymer paper-based batteries. Nano Lett 2009;9:3635e9.

[41] Wang Z, Tammela P, Zhang P, Strømme M, Nyholm L. Efficient high activemass paper-based energy-storage devices containing free-standing additive-less polypyrroleenanocellulose electrodes. J Mater Chem A 2014;2:7711e6.

[42] Pan R, Xu X, Sun R, Wang Z, Lindh J, Edstrom K, Stromme M, Nyholm L.Nanocellulose modified polyethylene separators for lithium metal batteries.Small 2018;14:1704371.

[43] Chen W, Yu H, Lee S-Y, Wei T, Li J, Fan Z. Nanocellulose: a promising nano-material for advanced electrochemical energy storage. Chem Soc Rev2018;47:2837e72.

[44] Wang Z, Pan R, Sun R, Edstr€om K, Strømme M, Nyholm L. Nanocellulosestructured paper-based lithium metal batteries. ACS Appl Energy Mater2018;1:4341e50.

[45] Liu X, Cao J, Li H, Li J, Jin Q, Ren K, Ji J. Mussel-Inspired polydopamine: abiocompatible and ultrastable coating for nanoparticles in vivo. ACS Nano2013;7:9384e95.

[46] Lee H, Dellatore SM, Miller WM, Messersmith PB. Mussel-inspired surfacechemistry for multifunctional coatings. Science 2007;318:426e30.

[47] Liu R, Mahurin SM, Li C, Unocic RR, Idrobo JC, Gao H, Pennycook SJ, Dai S.Dopamine as a carbon source: the controlled synthesis of hollow carbonspheres and yolk-structured carbon nanocomposites. Angew Chem Int Ed2011;50:6799e802.

[48] Barclay TG, Hegab HM, Clarke SR, Ginic-Markovic M. Versatile surface modi-fication using polydopamine and related polycatecholamines: chemistry,structure, and applications. Adv. Mater. Interfaces 2017;4:1601192.

[49] Liu T, Kim KC, Lee B, Chen Z, Noda S, Jang SS, Lee SW. Self-polymerizeddopamine as an organic cathode for Li- and Na-ion batteries. Energy EnvironSci 2017;10:205e15.

[50] Ek R, Gustafsson C, Nutt A, Iversen T, Nystr€om C. Cellulose powder fromCladophora sp. algae. J Mol Recogn 1998;11:263e5.

[51] Mihranyan A. Cellulose from cladophorales green algae: from environmentalproblem to high-tech composite materials. J Appl Polym Sci 2011;119:2449e60.


[52] Liu Q, Wang N, Caro J, Huang A. Bio-inspired polydopamine: a versatile andpowerful platform for covalent synthesis of molecular sieve membranes. J AmChem Soc 2013;135:17679e82.

[53] Zhou J, Wang P, Wang C, Goh YT, Fang Z, Messersmith PB, Duan H. Versatilecoreeshell Nanoparticle@Metaleorganic framework nanohybrids: exploitingmussel-inspired polydopamine for tailored structural integration. ACS Nano2015;9:6951e60.

[54] Liu M, Zeng G, Wang K, Wan Q, Tao L, Zhang X, Wei Y. Recent developments inpolydopamine: an emerging soft matter for surface modification andbiomedical applications. Nanoscale 2016;8:16819e40.

[55] Pan RJ, Wang ZH, Sun R, Lindh J, Edstrom K, Stromme M, Nyholm L. Thicknessdifference induced pore structure variations in cellulosic separators forlithium-ion batteries. Cellulose 2017;24:2903e11.

[56] Jacques E, Hellqvist Kjell M, Zenkert D, Lindbergh G, Behm M. Expansion ofcarbon fibres induced by lithium intercalation for structural electrode appli-cations. Carbon 2013;59:246e54.

[57] Takami N, Satoh A, Ohsaki T, Kanda M. Lithium insertion and extraction forhigh-capacity disordered carbons with large hysteresis. Electrochim Acta1997;42:2537e43.

[58] Zheng T, McKinnon WR, Dahn JR. Hysteresis during lithium insertion inhydrogen-containing carbons. J Electrochem Soc 1996;143:2137e45.

[59] Wang Z, Xiong X, Qie L, Huang Y. High-performance lithium storage innitrogen-enriched carbon nanofiber webs derived from polypyrrole. Electro-chim Acta 2013;106:320e6.

[60] Yoshio M, Wang H, Fukuda K, Hara Y, Adachi Y. Effect of carbon coating onelectrochemical performance of treated natural graphite as lithium-ion bat-tery anode material. J Electrochem Soc 2000;147:1245e50.

[61] Rouquerol J, Avnir D, Fairbridge C, Everett D, Haynes J, Pernicone N, Ramsay J,Sing K, Unger K. Recommendations for the characterization of porous solids(Technical Report). Pure Appl Chem 1994;66:1739e58.

Ruijun Pan, who is currently a PhD student in theDepartment of Chemistry-Ångstr€om at Uppsala University,has a Bachelor degree in Materials Science and Engineeringfrom Zhejiang University and Master degrees in Sustain-able Energy Technology from Technical University ofEindhoven and Royal Institute of Technology. His PhDthesis has been focused on the development of cellulose-based functional separators for lithium batteries.

separators for lithium

Dr. Zhaohui Wang received his B.E. (2007) and M.E. (2009)and Ph.D. degree (2012) from Huazhong University ofScience and Technology, China. He works as a researcher inUppsala University since 2013. His research focus on cel-lulose and conducting polymers, with application in thefield of electrochemistry and energy storage.

-ion batteries, Journal of Materiomics, https://






































































































































journal of materiomics - diva portaluu.diva-portal.org/smash/get/diva2:1272288/fulltext01.pdf ·...

Documents