Scalable Synthesis of Microsized, Nanocrystalline Zn0.9Fe0.1O‐C Secondary Particles and Their Use in Zn0.9Fe0.1O‐C/LiNi0.5Mn1.5O4 Lithium‐Ion Full Cells

Abstract Conversion/alloying materials (CAMs) are a potential alternative to graphite as Li‐ion anodes, especially for high‐power performance. The so far most investigated CAM is carbon‐coated Zn0.9Fe0.1O, which provides very high specific capacity of more than 900 mAh g−1 and good rate capability. Especially for the latter the optimal particle size is in the nanometer regime. However, this leads to limited electrode packing densities and safety issues in large‐scale handling and processing. Herein, a new synthesis route including three spray‐drying steps that results in the formation of microsized, spherical secondary particles is reported. The resulting particles with sizes of 10–15 μm are composed of carbon‐coated Zn0.9Fe0.1O nanocrystals with an average diameter of approximately 30–40 nm. The carbon coating ensures fast electron transport in the secondary particles and, thus, high rate capability of the resulting electrodes. Coupling partially prelithiated, carbon‐coated Zn0.9Fe0.1O anodes with LiNi0.5Mn1.5O4 cathodes results in cobalt‐free Li‐ion cells delivering a specific energy of up to 284 Wh kg−1 (at 1 C rate) and power of 1105 W kg−1 (at 3 C) with remarkable energy efficiency (>93 % at 1 C and 91.8 % at 3 C).


Introduction
Since the first commercialization by Sony in 1991, the market for lithium-ion batteries (LIBs) has been growing beyond expectations. [1]10] Thisi ncreasing diversity of potential applications, however,a lso leads to ag reater variety of required characteristics.For example, the use in hybrid electric vehiclesr equires high powerd ensity while maintaining high safety and energy density as well as low cost. [11]In addition, because of the rapid increaseo fp roduction associated with the rapidlyg rowing electrification of the automotive sector, [8,12] LIBs have to become more sustainable.This meanst hat critical elements such as cobalt must be omitted, which concerns basically the cathode.[14][15] Accordingly, cobalt-free cathode materials such as the high-voltage spinel LiNi 0.5 Mn 1.5 O 4 (LNMO), initially reportedb yA mine et al. [16] and Zhong et al., [17] have attracted increasing interest.In fact, LNMO additionally provides excellent fast charging characteristics due to the 3D Li + diffusion pathways in the spinel structure, while the high delithiation/lithiation potential of approximately 4.7 Ve nsures high energy and power densities. [18]evertheless, the rate capability of the final full cell is commonly not determined by the cathode, but rather by the sluggish lithiation kinetics of the graphite anode. [19][22][23][24] Recently,athird class of compounds, namely,s o-called conversion/alloyingm aterials (CAMs), [25] has attracted increasing attention.C AMs combine both reactionmechanismsinasinglematerialbyins itu formation of nanograins of an alloying elementa nd ap ercolating Conversion/alloying materials (CAMs)are apotentialalternative to graphite as Li-ion anodes, especially for high-power performance.The so far most investigated CAM is carbon-coated Zn 0.9 Fe 0.1 O, which provides very high specific capacity of more than 900 mAh g À1 and good rate capability.E specially for the latter the optimal particles ize is in the nanometer regime.However,t his leads to limited electrode packing densities and safety issues in large-scale handling and processing.Herein, a new synthesis route including three spray-drying steps that results in the formation of microsized,s pherical secondary parti-cles is reported.The resulting particles with sizes of 10-15 mm are composed of carbon-coated Zn 0.9 Fe 0.1 On anocrystals with an average diameter of approximately 30-40nm.The carbon coating ensures fast electron transport in the secondary particles and, thus, high rate capability of the resulting electrodes.Couplingp artially prelithiated, carbon-coated Zn 0.9 Fe 0.1 O anodesw ith LiNi 0.5 Mn 1.5 O 4 cathodes resultsi nc obalt-free Li-ion cells delivering as pecific energy of up to 284 Wh kg À1 (at 1C rate) and power of 1105 Wkg À1 (at 3C)w ith remarkableenergy efficiency (> 93 %at1Cand 91.8 %at3C).
conductive network of transition metal nanograins on lithiation;t he latter even allow for reversible cycling of the simultaneously formed Li 2 Om atrix. [25]One of the most investigated CAMs is Zn 0.9 Fe 0.1 O. [26][27][28][29][30][31] Besides being composed of environmentally friendly and abundant elements, it offers ah igh specific capacity of 966 mAh g À1 and very good rate capability.F or both advantageous properties, however,t he use of nanosized particles is essential, whichi sa no bstacle for the realization of high-density electrodes and, thus, suitable volumetric energy densities and their handling on an industrial scale, that is, their potentialapplication in commerciald evices.
Herein, we report an ew,s calables ynthesis route involving three spray-drying steps that allows for the preparation of microsized but nanocrystalline carbon-coated Zn 0.9 Fe 0.1 O (Zn 0.9 Fe 0.1 O-C) secondary particles.The nanometric crystallite size of the primary particles ensures good electrochemical performance, while the large secondary particles ize of about 10-15 mmf acilitates handling and processing.The subsequent combination of this materiala sn egative electrode with an LNMO-based positivee lectrode enabled the first full cellso f this kind showing potentially high energy efficiency and suitable specific energy.

Results and Discussion
Synthesis and characterization of carbon-coated Zn 0.9 Fe 0.1 O Carbon-coated Zn 0.9 Fe 0.1 O( Zn 0.9 Fe 0.1 O-C) was synthesized by a newly developed and readily scalables ynthesis method, as summarized in Figure 1.Firstly,Z n 0.9 Fe 0.1 On anoparticles were synthesized by spray drying of an aqueous solution of zinc(II) acetate andi ron(II) gluconate precursors (9:1 molar ratio),b oth of whicha re rather cost-efficient and environmentally friendly chemicals.The obtained compound was calcined in ab ox fur-nace at 450 8Cf or 3h to give phase-pure, wurtzite-structured Zn 0.9 Fe 0.1 On anoparticles (Figure 2a).These have an average particles ize of about3 0nm( Figure2b) and BET surface area of 33.5 m 2 g À1 .C ompared with the laboratory-scale synthesis reported earlier, [28] this corresponds to aslightincrease in particle size (formerly, < 20 nm) and decrease in BET surface area (formerly, 90 m 2 g À1 ).While these slight differences are certainly related to the differents ynthetic method, the choice of the precursors deserves brief reconsideration.In fact, the laboratory-scales ynthesis involving the gluconate salts of both metals [26] leads to as ignificant volumee xpansion on synthesis due to the formationo favoluminous "foam" when the temperaturei si ncreased.For larger batches, this is rather hard to handle.Using the acetate salts insteadc an successfully address this issue, but results in the formation of impurity phases in the final product, so that it was not possible to reach homogeneous doping of Fe in the ZnO lattice (Figure S1 in the Supporting Information).In contrast, the combination of zinc acetate and iron gluconate allowed for the synthesis of ap hasepure material while suppressing the extensive foaming.Similarly,t he carbon coating procedure had to be adapted.In fact, the use of sucrosea sc arbon precursor turned out to be challenging.The wet grindingo ft he activem aterial with sucrose leads to the formation of as oufflØ-like foam, and the rather pronounced hygroscopic nature of sucrose in combination with its relatively low glass transition temperature limit the potentialp rocessing window for the subsequent spray-drying step. [32]These issues could be tackled by replacings ucrose with b-lactose and dispersing the active material without an additional grinding step in the ethanolic solution of b-lactose, followed by spray drying the resulting dispersion.After calcination of the dried dispersion at5 00 8Cu nder argon atmosphere, the powder was ground and subsequently granulated in an additional spray-drying step.The XRD pattern of the resulting Zn 0.9 Fe 0.1 O-C is shown in Figure 3a.N oa dditional reflections are observed, and this confirmed that no phase impurities were introduced duringt he carbon-coatingp rocess, while the width of the reflections is comparable to that of the uncoated Zn 0.9 Fe 0.1 O, that is, the particle size did not increased uring the additional heat treatments.T he total carbon coating content was determined by thermogravimetric analysis( TGA) to be approximately 13 wt % (Figure S2).The morphology of the granulated materialw as studied by SEM (Figure 3b), which revealed that the Zn 0.9 Fe 0.1 O-C powder consists of spherical secondary particles with ad iametero fa pproximately 10-15 mm.These relatively large particles are composed of nanometric primary particles in the range of about3 0-40 nm on average, as is apparent from the SEM image shown in Figure 3c,w hich further confirms that the initial particlesp rior to carbon coating were well maintained.The crosss ection of as ingle granule is shown in Figure 3d,w hich shows that these microsized secondary particles are composed of denselyp acked agglomerates of the nanocrystalline primary particles with as ize of less than 1 mm.The pores between these agglomerates may facilitate electrolyte penetration into the secondary particles and, thus, favor the discharge/charge kinetics, thought he determined true density of 4.1gcm À3 is, as ac onsequence (in the case of inaccessible pores) and as ar esult of the carbon content ( % 13 wt %), somewhat lower than the theoretical value for pure ZnO (5.6 gcm À3 ).The energy dispersive X-ray (EDX) spectroscopic mapping for Zn, O, Fe, and C( Figure 3e)a sw ell as the longitudinal (normalized) elemental analysis( Figure 3f) along the horizontal white line shown in Figure 3d reveal that all elements are homogeneously distributed in these secondary particles.Considering carbon, this means that the single nanocrystalsa re also electronically well connected, and this suggests that ions and electrons can move rapidlyf rom the outer shell into the core of the microsized secondary particles,w hich is essential for achieving high power.T oo btain more detailed information aboutt he morphology and structure of the carbon-coated Zn 0.9 Fe 0.1 Om aterial, we performed HRTEM (Figure 4). Figure 4a shows two micrographs at the outer edge of as econdary particle for studying the size of the primary nanocrystals.Globally,as ize distribution of about 15-80 nm is observed, with the majority of the particles having ad iameter of approximately 30-40 nm, in line with the SEM observation.In Figure 4b two additional micrographs at higherr esolution revealt hat the primary particlesa re highly crystalline with fringes of,f or example, about 0.25 nm for the interlayer spacing of the (101) planes [33] (highlighted in yellow).For the HRTEMi mage in Figure4c, the focus was on studying the distribution of the carbon coating att he local scale.It is apparent that the amorphous carbon (in line with the absence of any additional carbon-related reflection in Figure 3a)t horoughly interconnects the single primary particles at the corresponding interfaces and largely covers the surfaceo ft he primaryp articles with al ayer of severaln anometers (exemplarily illustrated by the yellow arrows).It is also observed, however, that some primary particles are not fully covered by carbon at the outer surfaceo ft he secondary particle.Nonetheless, this does not hamper the electron transport in the secondary particles.

Electrochemical characterization
For the electrochemical characterization of Zn 0.9 Fe 0.1 O-C, halfcells assembled with lithium metal as counter electrode were subjected to galvanostaticc ycling (Figure5).For this basic characterization, the whole voltage range of the electrochemical activity of Zn 0.9 Fe 0.1 Ow as explored.Firstly,w ee valuated the constant-current cycling at ar elatively low specific current (100 mA g À1 )after one formation cycle at 50 mA g À1 (Figure 5a).The coulombic efficiency in the first cycle is approximately 70 %( see also Figure 5b)b ut increases to about9 7-99 %l ater, depending on the Cr ate (Figure 5a).The reversible specific capacity decreases in the initial ten cycles (Figure 5a andc ).However, on furtherc ycling it stabilizes at about8 50 mAh g À1 and even tends to slightly increase later on (Figure5aa nd d).Such atrend, observed earlier for other conversion and conversion/alloying materials, has been assigned to the quasireversible formation of the solid electrolyte interphase (SEI)l ayer. [34][37][38][39] We note, however,t hat this increase is only marginal compared to earlier studies, [36][37][38] and this suggests that the carbonc oating better stabilizes the active mate- rial/electrolyte interface, that is, suppressest he dissolution and reformation of the SEI.Thisq uasireversible SEI formation can be effectively inhibited by limiting the upperc utoff voltage to 2.0 Vo rl ess, as confirmedb yarecents tudy involving in situ microcalorimetry. [40]he rate capability of the Zn 0.9 Fe 0.1 O-C electrode was investigated by subjecting the cells to discharge/charge rates ranging from C/10 to 10C (i.e.,s pecific currents ranging from 100 to 10 000 mA g À1 ;F igure 5e and f).Beforei ncreasing the discharge/charge rate, the cellswere cycled for ten cycles at C/10, that is, until the coulombic efficiency hads tabilizeda ta pproximately 97 %.As the Cr ate increasesf rom 100 to 200, 500, 1000, 2000, 5000, and 10 000 mA g À1 ,t he capacity decreases from 810 to 770, 710, 650, 410, and 240 mAh g À1 ,r espectively, indicating very good rate capability of the microsized, nanocrystalline Zn 0.9 Fe 0.1 O-C.Thisi sa lso reflectedb yt he relatively small increase in polarizationc onsidering the currents applied (Figure 5f).After this C-rate test, the specificc urrent was decreaseda gaint o2 00 mA g À1 (C/5), which resulted in ac apacity of approximately 770 mAh g À1 ,t hat is, the same value as beforet he C-rate test at this specific current.This confirms the good reversibility of the delithiation/lithiation mechanism of Zn 0.9 Fe 0.1 O-C.
Generally,t heser esults clearly exceed the rate capability data reported earlier for the materials derived from laboratoryscale synthesis. [26,28] or instance, someo fu sh ave previously reported specific capacities of approximately 450, 300,a nd 110mAh g À1 at specific currents of 1000, 2000, and 5000 mA g À1 ,r espectively, [28] which are substantially lower than the capacities presented herein, that is, ca.650, 410, and 240 mAh g À1 ,r espectively.T his brief comparison further highlights that the newly developed, scaled-up synthesis does not have any negative, buti nsteadahighlya dvantageous impact on the material performance.Zn 0.9 Fe 0.1 O-C/LiNi 0.5 Mn 1. 5

O 4 lithium-ion cells
To demonstrate the potential of Zn 0.9 Fe 0.1 O-C as alternative anode material for high-power Li-ion batteries, we combined Zn 0.9 Fe 0.1 O-C negative electrodes with high-voltage LNMO positive electrodes.To the best of our knowledge,t his is the first Li-ion cell of such ak ind.As the first cycle coulombic efficiency of Zn 0.9 Fe 0.1 O-C still deserves furtheri mprovement, the negative electrodes were first prelithiated.To fine-tune the operational potentialr ange of the anode three differently prelithiated sets of electrodes were used, followingapreviouss tudy. [41]or the first one, the anodesw ere lithiated and completely delithiated prior to full-cell assembly (Zn 0.9 Fe 0.1 O-deLi;F igure6ac).For the second one, the anodes were partially lithiated with as pecific capacity of 300 mAh g À1 (Zn 0.9 Fe 0.1 O-300;F igure 6df), and for the third one the anodesw ere partially lithiated with as pecific capacity of 600 mAh g À1 (Zn 0.9 Fe 0.1 O-600;F igure 6g-i).Note that the specific capacities given in Figure 6 refer to the mass of both active materials, that is, the negative and positive electrodes.For easier comparison of the performance of the cathode and anode individually,t he same plot is provided in Figures S3 and S4, in which the specific capacities refer to the active-material mass loading of the LNMO cathodea nd the Zn 0.9 Fe 0.1 Oa node,r espectively.F or ag raphite/ LNMO cell with aN /P capacity ratio of 1.2, the theoretical specific capacity of the full cell would correspond to 80 mAh g À1 (assuming ar eversible specific capacity of 350 mAh g À1 for the graphite anode).In the Li-ion cells investigated herein, the P/N mass ratio was 2.36, 2.68, and 1.68, respectively for Zn 0.9 Fe 0.1 O-deLi, Zn 0.9 Fe 0.1 O-300, and Zn 0.9 Fe 0.1 O-600.Considering the substantially higherc apacity of the anode, this means that all cells were cathode-limited with N/P capacity ratios of 3.42, 1.95, and 1.42 for the full cells with Zn 0.9 Fe 0.1 O-deLi, Zn 0.9 Fe 0.1 O-300, and Zn 0.9 Fe 0.1 O-600 anodes, respectively.T hese values were calculated on the basis of the practically obtained capacities of these electrodes at the given discharge/charge rate, that is, 105 mA g À1 for LNMO and 850 mAh g À1 ,5 50 mAh g À1 ,a nd 250 mAh g À1 ,f or Zn 0.9 Fe 0.1 O-deLi, Zn 0.9 Fe 0.1 O-300,a nd Zn 0.9 Fe 0.1 O-600,r espectively.W hen cycled at 1C (147 mA g À1 ), all cellss howeda ni nitial increasei nc apacity over several cycles (Figure 6a,d,and g), which is related to acontinuous increase in capacity for the LNMO cathode, as is apparent from the discharge/charge profile evolution.The steady increase in capacityo ccurs along the high-voltage plateau (Figure 6b,e ,  h) originating from the Ni 2 + QNi 3 + QNi 4 + redox reaction. [32]This already-observed phenomenon has been assigned to the sluggish electrolyte wetting of the aqueous-processed LNMO cathode. [42]he full cell with Zn 0.9 Fe 0.1 O-deLi as negative electrode (Figure 6a)s hows an initial increase in capacity,f ollowed by slight but steady fading.T he anode reachesi mmediately the upper cutoff voltage of 3.0 V, while the lower cutoff of the cathode rapidly rises, so that eventually only the nickel redox couple at about 4.7 Vi su tilized (Figure6b).In combination with the rather low coulombic efficiency ( % 99 %) during the initial 15 cycles,t his behavior indicates ongoing lithium loss.The discharge/charge profile for the 25th cycle (Figure 6c)r eveals that the anode does not reach its lower cutoff voltage and delithiation/lithiation occurse ssentially in the regime of the conversion reaction. [26,31,40] Tis results in ar ather low energy efficiency (EE) of approximately 78.2 %, an averagef ull-cell voltage of 3.0 V, and as pecific energy of 157 Wh kg À1 for the 25th cycle.Although such an EE is rather low compared to lithium-ion cells with ag raphite anode, it is significantly highert han that of only approximately 62 %r eported earlier for (theoretical) ZnFe 2 O 4 /LiFePO 4 full cells. [43]o avoid the energy loss associated with the anode operating at high voltages, the prelithiated Zn 0.9 Fe 0.1 O-300 electrode (300 mAh g À1 )w as employedi nt he full cell.This "lithium reservoir" allows for substantially higherc apacities, which stabilize at approximately 80 mAh g À1 (i.e.,t he same specific full-cell capacity as foragraphite/LNMOl ithium-ion cell), and cycling stability (Figure 6d).The capacity-limiting LNMO cathode cycles stably within the set cutoff potentials, so that the characteristic feature of the Mn 3 + /4 + redox reaction at approximately 4.1 Vi s well maintained (Figure 6e).In fact, recalculating the specific capacityf or the LNMO cathode gave av alue of approximately 110mAh g LNMO À1 ,t hat is, the full capacity of the cathode is used.Similarly,t he operational voltage range of the anode is sub- stantially extended to lower potentials compared with the full cell based on Zn 0.9 Fe 0.1 O-deLi, without hitting the upper cutoff potentialo f3 .0V. Nevertheless,as light increasei so bserved on cycling with av ery particular feature occurring after about ten cycles.Towards the end of the discharges tep, av oltage "bump" is recorded, which gets more pronouncedo nf urther cycling.[46] This feature, which is considered to have ar ather negative impact on the long-term cycling performance and stability of the SEI at the anode, [47][48][49] has not been observed for the Zn 0.9 Fe 0.1 O-deLi/LNMO full cell, since the manganese redox process at the cathode did not occur extensively owing to the fast capacityf ading.Accordingly,t he LNMO cathoder equires further improvement foro peration in lithium-ion cells.Nonetheless, the partial prelithiation allowsf or as ignificant increase of the average cell voltage (3.3 V), specific energy (262 Wh kg À1 ), and EE (83.2 %), as exemplarily determined fort he 25th cycle at 1C (Figure 6f).In fact, the shift of the operational potential of the anode to lower values, accompanied by an increased share of the alloying contribution,r esults in an appreciable decrease of the voltage hysteresis and, thus, ahigherEE.)having anodes with different degrees of prelithiation.a-c)Zn 0.9 Fe 0.1 O-deLi, d-f) Zn 0.9 Fe 0.1 O-300,and g-i)Zn 0.9 Fe 0.1 O-600.For each full cell the following plots are shown(from left to the right): plot of the specific capacity versus cycle number,t he corresponding separate discharge/charge profilesf or all cycles for the full cell(green),t he anode (blue), and the cathode (red),a nd the slightly modified plot of the discharge/chargep rofile for the 25th cycle,h ighlighting the voltage hysteresis between the charge and discharge processes.Note that the specificc apacitiesa re basedo nt he sum of the anode andcathode active materials.Priort oc ycling at 1C,af ormation cyclea tC /10 was applied to each cell.The potentialsoft he LNMOc athodes and the Zn 0.9 Fe 0.1 Oanodewere limitedt o3 .5-4.8 and 0.01-3.0V, respectively.
To explore the effect of more extensive prelithiation, we increasedt he lithium reservoir in the anode to 600 mAh g À1 (i.e., Zn 0.9 Fe 0.1 O-600).Just as in the previousc ase, the capacity initially rises to approximately 70 mAh g À1 after 25 cycles (Figure 6g).Althought his corresponds again to as pecific capacity of approximately 110mAh g À1 for the cathode, that is, its complete utilization, the overall value is lower due to the larger anode and its lower (remaining) capacity after the prelithiation.Also, the overall cell capacity fades rather rapidly on cycling, even faster than that of the Zn 0.9 Fe 0.1 O-deLi/LNMO full cell (Figure 6a).Considering the large lithium reservoir,i ta ppearsu nlikely that this fading is related to lithium loss.In fact, the voltage profile of the cathode is well maintained in its shape (Figure 6h).Instead, the shape of the anode discharge/charge profile changes to ag reater extent.I np articular,t he feature assigned to manganese reduction and reoxidation at the anode is more pronounced in this case, and this highlights its detrimental effect on the fullcell performance.Indeed,t he manganese concentration on the Zn 0.9 Fe 0.1 O-600 anode is rather high, as confirmed by ex situ SEM/EDX analysis (Figure 7).Nonetheless, for the exemplary 25th cycle, the Zn 0.9 Fe 0.1 O-600/LNMOl ithium-ion cell providesr emarkable average voltage (4.1 V), specific energy (284 Wh kg À1 ,i .e.,a lmost double the specific energy of the Zn 0.9 Fe 0.1 O-deLi/LNMO cell), specific power( 375 Wkg À1 ), and EE (> 93 %; Figure 6i), av alue that is comparable to those of state of the art graphite-based Li-ion cells. [43]Since these values again refer only to the two activem aterials at the negative and positive electrodes, so that for direct comparison with commercial cells also the inactivec omponents would have to be considered (commonly,e xtensive optimization is done prior to any commercialization).Nevertheless, the EE is not affected by the presence and any optimization of the inactivec omponents (apart from polarization effects)a nd the specific energy reported herein is rather comparable to that of al ittle more than 300 Wh kg À1 recently reported for ag raphite/LNMO laboratory-scale full cell, albeit at al ower discharge/charge rate of C/3. [50]As ummary of the results obtained for the three full cells is provided in Ta ble 1.
Ah igh EE is, in fact, not only important with regard to general aspects such as sustainability and cost, but especiallyr elevant for high-power batteries,s ince energy inefficiency is largely released as heat, [40] whichm ight be an issue in practical application.H ence, we also subjected LNMO/Zn 0.9 Fe 0.1 O-600 full cells to elevated discharge/charge rate (3 C) to study the impact of such increased current on the EE, while simultaneously evaluating the general applicabilityo fs uch lithium-ion cells for high-power devices (Figure 8).Generally,t he cell shows as imilar behavior to that cycled at 1C (Figure 6g), that is, an initial increaseu pt oa bout 58 mAh g À1 at the 70th cycle (corresponding to 92 mAh g À1 for the LNMO cathode only) and subsequentr apid decrease.W hile these results further highlight the need for an optimized cathode to suppress dissolution of manganese andi ts subsequent redox activity at the anode,t he specific energy (230 Wh kg À1 )a nd power (1105 Wkg À1 ), EE (91.8 %), and average discharge voltage (4.0 V) make this lithium-ion cell chemistry very suitable for high-power applications.

Conclusions
We have described an ew,s caled-up synthesis route for the preparation of carbon-coated Zn 0.9 Fe 0.1 Oc omposed of nanocrystalline particles agglomeratedi nto microsized secondary particles.The material provides very good electrochemical performance in half-cells with ar eversible capacity of approximately 850 mAh g À1 at C/10 and high rate capability.The investigation of different degrees of prelithiation in Zn 0.9 Fe 0.1 O/ LNMO full cells revealed the beneficial effect of limiting the operational potentialo ft he anode to the alloying-dominated regime while granting al ithium reservoir.A saresult,t hese cells showeds pecific energies of up to 284 Wh kg À1 at 1C and 230 Wh kg À1 at 3C,c orresponding to specific powers of 375 and 1105 Wkg À1 ,r espectively.R emarkably,t he energy efficiencies at such discharge/charge rates are as high as > 93 %( 1C)

Experimental Section Material synthesis
To obtain microsized nanocrystalline Zn 0.9 Fe 0.1 O-C, at wo-step synthesis was used.First, the zinc(II) acetate dihydrate (Alfa Aesar) and iron(II) d-gluconate dihydrate (Aldrich) precursors were dissolved in water and the solution was spray dried with aG EA Niro Mobile Minor spray dryer to synthesize Zn 0.9 Fe 0.1 On anoparticles.Subsequently,t he powder was calcined at 450 8Cf or 3h(VMK-1400, Linn High Therm) and afterwards ground by planetary ball milling (Pulverisette 5, Fritsch) for 24 hb yu sing yttria-stabilized zirconia beads.The carbon coating was achieved by spray drying ad ispersion of the Zn 0.9 Fe 0.1 On anoparticles in an aqueous solution of blactose, followed by calcination of the resulting powder at 500 8C for 4hunder an argon atmosphere (VMK-135-S, Linn High Therm).Finally,t he carbon-coated Zn 0.9 Fe 0.1 On anoparticles (Zn 0.9 Fe 0.1 O-C) were ground again by planetary ball milling for 2hand granulated by spray drying.

Structural and Morphological Characterization
The powder properties of the synthesized Zn 0.9 Fe 0.1 Ow ere investigated by powder XRD (D5005, Siemens), field-emission SEM (Supra 55, Zeiss), and nitrogen physisorption (Gemini VII 2390, Micromeritics).The XRD measurements were performed with Cu Ka radiation in a2 q range of 15-808,a nd SEM images were obtained at an accelerating voltage of 10 kV.T he specific surface area was calculated according to the BET theory.T he structure of Zn 0.9 Fe 0.1 O-C was studied by XRD with aB ruker D8 Advance diffractometer (Cu Ka radiation, l = 0.154 nm) in the 2q range of 20-908.S EM was conducted with aZ eiss Crossbeam 340 field-emission electron microscope, equipped with an EDX spectrometer (Oxford Instruments X-MaxN, 50 mm 2 ,1 5kV) and aC apella gallium-focused ion beam (FIB).For the ex situ EDX measurements, the cycled electrodes were recovered in an argon-filled glove box, carefully rinsed with dimethyl

Electrodepreparation
For electrode preparation, Zn 0.9 Fe 0.1 O-C and carbon black (Super C65, Imerys) were added to a1 .25 wt %s olution of sodium carboxymethyl cellulose (CMC, Dow WolffC ellulosics) in deionized water.T he composition of the dry materials in the slurry was 75 wt %Z n 0.9 Fe 0.1 O-C, 20 wt %c arbon black, and 5wt% CMC.The slurry was mixed by planetary ball milling (Pulverisette 4, Fritsch) for 2h.T he homogenized slurry was then cast on dendritic copper foil (Schlenk) by using al aboratory doctor blade with aw et-film thickness between 120 and 200 mma nd subsequently dried at 80 8Cf or 5min and 12 ha tr oom temperature.Disk electrodes (12 mm diameter) where punched and dried for 12 ha t1 20 8C under vacuum.The LNMO cathodes for full-cell assembly where prepared as reported by Kuenzel et al. [42] Electrochemical characterization The electrochemical characterization was performed in three-electrode Swagelok-type cells, assembled in an argon-filled glove box (MBraun, Germany;oxygen and water content < 0.1 ppm).As separator,asheet of glass fiber fleece (Whatman, GFD), soaked with a 1 m solution of LiPF 6 in am ixture of ethylene carbonate (EC) and diethyl carbonate (for the half-cell tests) or EC and DMC (for the full-cell tests with the LNMO cathodes) was used.In the half-cells battery-grade Li metal (Honjo) served as both counter and reference electrodes.The specific capacities provided herein are based on the mass of the active material including the carbon coating.For the full-cell tests employing lithium metal as quasireference electrode, the Zn 0.9 Fe 0.1 O-C anodes were galvanostatically precycled for ten cycles at C/10 (i.e.,0 .1 Ag À1 )i nt he potential range of 0.01-3.0Vv s.Li/Li + .E ventually,t he electrodes were partially lithiated, as indicated in the text.Subsequently,t he cells were disassembled under argon, and full cells were assembled from such anodes and fresh electrolyte and separator.

Figure 2 .
Figure 2. a) XRD pattern and b) SEM image of Zn 0.9 Fe 0.1 On anoparticles,a s obtained by step IofF igure 1.The PDF reference for hexagonal wurtzitestructured ZnO (PDF 01-071-6424)isp rovided at the bottom of a).

Figure 3 .
Figure 3. a) XRD pattern of the Zn 0.9 Fe 0.1 O-C powder.The PDF reference for hexagonal wurtzite-structuredZ nO (PDF 01-071-6424)i sp rovided at the bottom.b) SEM image of the finallyo btainedZ n 0.9 Fe 0.1 O-C powder.c)SEMimage of the cross section of as econdaryparticle at highmagnification to illustratethe size and morphologyo ft he primary nanoparticles.d)Cross section of asingle secondaryparticlea tl ower magnification.e)EDXmappingf or Zn (orange), oxygen (red), iron (blue), and carbon (green)a nd f) the normalized concentrationoft hesee lements along the horizontal white line in d).

Figure 4 .
Figure 4. (HR)TEM analysis of the Zn 0.9 Fe 0.1 O-C powderatd ifferent magnifications with afocusona)the analysis of the particle size distribution of the primary particles, b) their crystallinity,a nd c) the distribution of the amorphous carbon coating withint he secondaryparticles andatt heir surface.

Figure 5 .
Figure 5. a) Plot of capacity versus cycle numberf or Zn 0.9 Fe 0.1 O-C/Li half-cells subjected to as pecific current of 100 mA g À1 after the first formationc ycle at 50 mA g À1 (cutoff voltages:0.01 and 3.0 V). b-d)The corresponding discharge/chargep rofiles for b) the first cycle, c) cycles 2-10, and d) cycles 11-60.e) Plot of capacity versus cycle number for Zn 0.9 Fe 0.1 O-C/Li half-cellss ubjected to elevated discharge/charge rates ranging from C/10 to 10C (i.e.,specific currents of 100 to 10 000 mA g À1 ;c ut-off voltages:0.01 and 3.0 V) and f) the corresponding discharge/charge profiles.All specific capacities refer to the mass of the active material including the carbonaceous coating.

Figure 6 .
Figure 6.Galvanostatic cycling of Zn 0.9 Fe 0.1 O/LNMO full cells at 1C (147 mA g LNMO À1 )having anodes with different degrees of prelithiation.a-c)Zn 0.9 Fe 0.1 O-deLi, d-f) Zn 0.9 Fe 0.1 O-300,and g-i)Zn 0.9 Fe 0.1 O-600.For each full cell the following plots are shown(from left to the right): plot of the specific capacity versus cycle number,t he corresponding separate discharge/charge profilesf or all cycles for the full cell(green),t he anode (blue), and the cathode (red),a nd the slightly modified plot of the discharge/chargep rofile for the 25th cycle,h ighlighting the voltage hysteresis between the charge and discharge processes.Note that the specificc apacitiesa re basedo nt he sum of the anode andcathode active materials.Priort oc ycling at 1C,af ormation cyclea tC /10 was applied to each cell.The potentialsoft he LNMOc athodes and the Zn 0.9 Fe 0.1 Oanodewere limitedt o3 .5-4.8 and 0.01-3.0V, respectively.
carbonate (DMC), and transferred to the SEM under argon atmosphere with aspecially designed transfer box (Sample Transfer Shuttle, SEMILAB).For the FIB treatment, currents of 1.5 nA and 50 pA at an acceleration voltage of 30 kV were chosen for milling and polishing, respectively.H RTEM images where recorded with aC scorrected high-resolution transmission electron microscope (FEI Titan, 80-300 kV) operated at acceleration voltages of 80 and 300 kV.T he weight of the carbon coating was determined by TGA (Model Q5000, TA Instruments) in the temperature range of 40-850 8Cu nder an oxygen atmosphere.The true density of the carbon-coated sample (Zn 0.9 Fe 0.1 O-C) was determined by utilizing an AccuPyc II 1340 gas pycnometer and helium as working gas.

Figure 8 .
Figure 8. Galvanostatic cycling of Zn 0.9 Fe 0.1 O-600/LNMO full cells at 3C (441 mA g LNMO À1 ).a) Plot of the specific capacity versus cyclenumber.b) The corresponding discharge/chargep rofiles for all cycles for the full cell(green)a nd separate electrodes (red and blue for the cathode and anode, respectively).c) Slightly modified plot of the discharge/chargep rofile for the 70th cycle.Notet hat the specific capacities are based on the sum of those of the anode and cathode active materials.Prior to cycling at 3C af ormation cycle at C/10 was applied to the cell.The potentials of the LNMO cathodesa nd the Zn 0.9 Fe 0.1 Oa node were limitedto3 .5-4.8 and 0.01-3.0V, respectively.

Table 1 .
Mass ratio, N/P ratio, specific energy, average discharge voltage, ande nergy efficiency of Zn 0.9 Fe 0.1 O/LNMO full cells with different degrees of prelithiation obtained from galvanostatic cycling at 1C (147 mA g À1 LNMO ).
and 91.8 %( 3C), which suggest that this cell chemistry is generally suitable for high-power applications if the manganese dissolution from the cathode can be suppressed.