Determining Li+-Coupled Redox Targeting Reaction Kinetics of

Jan 10, 2018 - The redox targeting reaction of Li+-storage materials with redox mediators is the key process in redox flow lithium batteries, a promis...
2 downloads 18 Views 751KB Size
Subscriber access provided by READING UNIV

Letter

Determining Li+-Coupled Redox Targeting Reaction Kinetics of Battery Materials with Scanning Electrochemical Microscopy Ruiting Yan, Jalal Ghilane, Kia Chai Phuah, Thuan Nguyen Pham Truong, Stefan Adams, Hyacinthe N. Randriamahazaka, and Qing Wang J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.7b03136 • Publication Date (Web): 10 Jan 2018 Downloaded from http://pubs.acs.org on January 11, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry Letters is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

Determining Li+-Coupled Redox Targeting Reaction Kinetics of Battery Materials with Scanning Electrochemical Microscopy Ruiting Yan1, Jalal Ghilane2, Kia Chai Phuah1, Thuan Nguyen Pham Truong2, Stefan Adams1, Hyacinthe Randriamahazaka2*, Qing Wang1* 1

Department of Materials Science and Engineering, Faculty of Engineering, National University

of Singapore, Singapore 117576 2

Université Paris Diderot, Sorbonne Paris Cité, ITODYS, UMR 7086 CNRS, SIELE group, 15

rue Jean Antoine de Baïf, 75013 Paris, France

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 19

ABSTRACT

Redox targeting reaction of Li+-storage materials with redox mediators is the key process in redox flow lithium batteries, a promising technology for next generation large-scale energy storage. The kinetics of the Li+-coupled heterogeneous charge transfer between the energy storage material and redox mediator dictates the performance of the device, while as a new type of charge transfer process it has been rarely studied. Here, scanning electrochemical microscopy (SECM) was employed for the first time to determine the interfacial charge transfer kinetics of LiFePO4/FePO4 upon delithiation and lithiation by a pair of redox shuttle molecules FcBr2+ and Fc. The effective rate constant  was determined to be around 3.70-6.57×10-3 cm/s for the two-way pseudo-first order reactions, which features a linear dependence on the composition of LiFePO4 validating the kinetic process of interfacial charge transfer rather than bulk solid diffusion. In addition, in conjunction with chronoamperometry measurement the SECM study disproves the conventional “shrinking-core” model for the delithiation of LiFePO4 and presents an intriguing way of probing the phase boundary propagations induced by interfacial redox reactions. This study demonstrates a reliable method for the kinetics of redox targeting reactions, and the results provide useful guidance to the optimization of redox targeting systems for largescale energy storage.

ACS Paragon Plus Environment

2

Page 3 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

TOC GRAPHICS

KEYWORDS: Redox Targeting, Charge Transfer, LiFePO4, Energy Storage, SECM

ACS Paragon Plus Environment

3

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 19

Interfacial charge transfer of lithium ion storage materials upon delithiation/lithiation has been fervently studied, as the kinetics of this process critically affects the power performance of Liion batteries. In 2006, Wang et al. proposed the use of “redox targeting reaction” to deliver energy through chemical oxidation/reduction of energy storage materials with redox shuttle molecules.1 This ushers in a novel way for efficiently charging/discharging poorly conductive materials. One example is LiFePO4, which in most previous kinetics studies needs to be mixed with sufficient carbon additives upon direct charging/discharging 2-4 while no carbon is required with redox targeting reactions. Recently, this concept has been applied in redox flow lithium batteries (RFLB),5-7 with more attention also given to the chemical delithiation/lithiation kinetics of LiFePO4.8 For instance, in a double-layer electrode system, LiFePO4 is oxidized/reduced by redox molecules which diffuse through the intermediate insulating layer and get regenerated at the electrode. The reaction rate between the redox species and LiFePO4 is derived from extracting the diffusion length of redox species inside the LiFePO4 layer.8 Despite the valuable kinetic information these trials have obtained, the experimental time scale constrained by the experimental setup and electrode configuration limits the detectable reaction rates to the relatively slow processes such as charge carrier (e- or Li+) transport or phase boundary propagation inside the material. It is thus important to explore alternative methods to delve into the kinetics of interfacial charge transfer of LiFePO4 upon delithiation/lithiation. Scanning electrochemical microscopy (SECM) has proved to be an effective tool in the study of finite electron transfer kinetics9 with expanding applications in measuring heterogeneous chemical reaction rates between redox species and material/chemical at various interfaces, such as electronically resistive film/solution interface,10 liquid/liquid interface,11-12 redox enzymes on solid supports13 and semiconductor/electrolyte interfaces,14-19 etc. Inspired by these studies,

ACS Paragon Plus Environment

4

Page 5 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

herein the interfacial charge transfer kinetics of the chemical lithiation/delithiation of LiFePO4/FePO4 induced by redox mediators, or redox targeting reactions was scrutinized with the feedback mode of SECM under unbiased substrate conditions. The working principle of SECM feedback mode involves the diffusion-controlled oxidation/reduction of redox mediator at the electrode tip and the reverse process at the substrate surface at a rate determined by interfacial charge transfer (Scheme 1). The recorded current at the tip is thus a combined effect of the tip-substrate distance (normalized by tip radius, /) and the rate of interfacial charge transfer between the redox species and substrate material.

Scheme 1. Reactions at SECM tip and substrate during (a) delithiation and (b) lithiation of LiFePO4/FePO4 by FcBr2+ and Fc under feedback mode.

In our study, we used a dense tablet made of carbon-free LiFePO4 powder as the substrate to eliminate the interference of carbon additive on the surface kinetics and to differentiate the measured charge transfer with that at electronically conductive but chemically inert substrate.20-21 To further investigate the sensitivity of reaction rate to active material surface concentration, the substrate composition was varied as “xLiFePO4/(1-x)FePO4” with x = 0, 0.25, 0.5, 0.75 and 1 by mixing LiFePO4 and FePO4 powders in the designated ratios. FcBr2+ and Fc, with each E1/2 at 3.55 V and 3.25 V vs. Li/Li+, were selected as the redox mediators to chemically delithiate and

ACS Paragon Plus Environment

5

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 19

lithiate LiFePO4/FePO4 (3.45 V, which corresponds to the Fermi level of LiFePO4), respectively. In the delithiation process, FcBr2+ will be reduced back to reduced state by hole injection in the LiFePO4 particles, accompanying the oxidation of Fe(II) to Fe(III) and the release of lithium ions, while the lithiation process of FePO4 is vice versa (See Fig. S4). The relatively large potential difference between the molecule and material (~100 mV for FcBr2+ and ~200 mV for Fc) suggests that both reactions are likely to follow an irreversible charge transfer kinetics on the substrate. Such an irreversible chemical delithiation/lithiation at the LiFePO4/FePO4 substrate, and charge transfer at the tip and substrate are illustrated in Scheme 1(a) and (b), with reactions shown below: Upon delithiation: FcBr − e → FcBr  (reaction at the tip) 

FcBr  + LiFePO  FcBr + FePO + Li (reaction at the substrate)

(1)

(2)

Upon lithiation: Fc  + e → Fc (reaction at the tip) 

Fc + Li + FePO → Fc  + LiFePO (reaction at the substrate)

(3)

(4)

In a typical SECM measurement, a Pt disk electrode of 5 µm in radius calibrated by voltammetry (see Fig. S5) was used as the SECM tip to approach the xLiFePO4/(1-x)FePO4 substrate. The electrolyte consists of 0.5 M lithium bis(trifluoromethane)sulfonimide (LiTFSI) in propylene carbonate (PC). For each fixed mediator and sample composition, the approach curve was

ACS Paragon Plus Environment

6

Page 7 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

repeated on 3-4 separate spots to check the reproducibility of the result (see Fig. S6). The normalized approach curves for both delithiation and lithiation at five different compositions are summarized in Figure 1 (a) and (b) (see the normalization steps in Supporting Information). The experimental curves do not have the characteristic of reversible or quasi-reversible substrate reaction models22 and can be fitted with irreversible substrate kinetics model when d/a < 2 (see the fitting process in Fig. S6 and Eq. S1-S5), confirming our previous hypothesis. For both delithiation/lithiation, there is a continuous decay of the positive feedback with decreasing x from 1 to 0 for the delithiation process and reversal for the lithiation process. A nearly negative feedback is observed when the active material ratio turns to zero, i.e., FcBr2+ to delithiate 100% FePO4 and Fc to lithiate 100% LiFePO4, which resembles the behavior of other highly resistive semiconductor substrates, such as Fc on p-Si(111)-C16H33 surface19 and FcMeOH on TCNQ crystal (σ =310-12 S/cm).23

ACS Paragon Plus Environment

7

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 19

Figure 1. Normalized SECM feedback approach curves (open squares) for a 5 µm Pt disk electrode towards xLiFePO4/(1-x)FePO4 substrate. (a) Delithiation of xLiFePO4/(1-x)FePO4 by FcBr2+ oxidation at x=1, 0.75, 0.5, 0.25, 0; (b) Lithiation of xLiFePO4/(1-x)FePO4 by Fc reduction at 1-x=1, 0.75, 0.5, 0.25, 0. Solid lines in the same color correspond to the calculated curves for the approach of a UME with RG value of 5 towards samples with kinetics of mediator recycling using in (a) normalized rate constants  =1.46, 1.20, 0.83, 0.48 and 0.02, respectively; and in (b)  =1.00, 0.78, 0.62, 0.38 and 0.07, respectively. (c) Delithiation of fresh LiFePO4 by FcBr2+ and a spot of LiFePO4 substrate right after 60 s transient CA experiment; (d) Lithiation of FePO4, LiFePO4 in LiTFSI/ PC and FePO4 in TBAPF6/PC electrolyte by Fc.

ACS Paragon Plus Environment

8

Page 9 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

In addition, it is observed that the positive feedback gradually weakens when multiple measurements were performed on the same spot. To quantify the effect, a chronoamperometry (CA) experiment was conducted for the delithiation process, with the tip electrode kept at a fixed distance from the substrate, typically 1.5L (L=d/a). A potential step was then applied to oxidize FcBr2 and the evolution of tip current was recorded as a function of time. The effect of CA pretreatment on the SECM experiment is shown in Figure 1 (c), where the positive feedback was considerably attenuated after 60 s of CA pretreatment. As the LiFePO4 component is subjected to delithiation during the CA process, the evolution of the feedback level implies that the recorded charge transfer kinetics is highly dependent on the concentration of active material at the substrate surface. In a further experiment, when the tip-generated Fc approached 100% FePO4 substrate in electrolyte with tetrabutylammonium hexafluorophosphate (TBAPF6) in place of LiTFSI as the supporting salt, the elimination of Li+ resulted in a negative feedback, with normalized approach curve almost overlapping that of Fc approaching 100% LiFePO4 substrate (Figure 1 (d)). This attests that the finite charge transfer processes detected in Figure 1 (b) be originated from the substrate lithiation process, where Li+ ions in electrolyte are inserted into the FePO4 phase. Table 1. Effective rate constants of interfacial charge transfer for the delithiation/lithiation of xLiFePO4/(1-x)FePO4 by FcBr2+ and Fc, respectively. Substrate composition, x 0 0.25 0.5 0.75 1

Delithiation average keff (10-3 cm/s) 0.09 1.93 3.36 4.93 6.57

Lithiation average keff (10-3 cm/s) 3.70 2.96 2.30 1.34 0.28

ACS Paragon Plus Environment

9

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 19

The kinetic parameters extracted by fitting the experimental approach curves with irreversible substrate kinetics model are summarized in Table 1. The effective rate constants  of 100% LiFePO4/FePO4 delithiated/lithiated by FcBr2+ and Fc is 6.5710-3 cm/s and 3.7010-3 cm/s, respectively. The Li+-coupled interfacial charge transfer process for delithiation is slightly faster than lithiation despite a relatively small driving force, presumably attributed to the higher energy required for Li+ desolvation than solvation process during the extraction/insertion process. As the active material ratio (x) in substrate decreases from 1 to 0, the average  declines from 6.5710-3 to 0.0910-3 cm/s for the delithiation process and from 3.7010-3 to 0.2810-3 cm/s for the lithiation process. The dependence of  on the substrate active material ratio is more clearly demonstrated in Figure 2. A quasi-linear relationship can be derived for both delithiation/lithiation processes, although the fittings did not strictly cross the zero point. This could be rationalized if the exposed surface area of LiFePO4 and FePO4 scales to the specific molar ratio (x) in the xLiFePO4/(1-x)FePO4 substrate. Since the interfacial reaction rate is proportional to the surface area of active material available, the derived  based on the feedback current is, in turn, proportional to x or (1-x) in xLiFePO4/(1-x)FePO4. This quasi-linear dependence of  on the active material ratio, along with the effect of transient CA pretreatment, thus concertedly confirms that the measured kinetics with this method is at the interfacial level, rather than those limited by the transport of electronic carriers inside the poorly conductive solid nanocrystals that could be very slow. This also explains why the measured kinetic values in this study are 3 orders of magnitude higher than that measured in the aforementioned double-layer electrode method, in which the measured rate constants (2.204.4010-6 cm/s) are independent of x in xLiFePO4/(1-x)FePO4.

ACS Paragon Plus Environment

10

Page 11 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

Figure 2. Average  of the chemical delithiation/lithiation of xLiFePO4/(1-x)FePO4 by FcBr2+ and Fc at different x. Solid lines represent the linear fits of the experimental data.

Note that the estimated maximum possible charge/discharge rate based on the double-layer electrode measurement has been excessively high to sustain a high rate operation of RFLB. An estimation based on the interfacial kinetics will undoubtedly indicate that the chemical delithiation/lithiation of LiFePO4/FePO4 here will not be a limiting factor in pursing high power density of RFLB. The significantly larger interfacial charge transfer kinetics also suggests that increasing surface-to-volume ratio of poorly conductive energy storage materials which shortens the charge transport distance could be a more effective way for the enhancement of overall power performance of RFLB. This could be achieved by modulating the morphology, porosity and loading of active materials, etc. Although the  measured here, which specifically corresponds to the charge transfer rates between the LiFePO4/FePO4 and the redox molecules used in this study, cannot to be directly copied to the chemical delithiation/lithiation of other LiFePO4 samples in other redox targeting

ACS Paragon Plus Environment

11

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 19

systems, they do provide a reliable range of interfacial reaction rate between LiFePO4/FePO4 and FcBr2+/Fc. To verify this assumption, we prepared another batch of carbon-free LiFePO4/FePO4 powders using hydrothermal method and repeated all the above feedback experiments. The  obtained from the home-made LiFePO4 is 3.9310-3 cm/s for the delithiation and 2.3510-3 cm/s for the lithiation process, respectively, and a linear dependence of  on the substrate composition x or (1-x) can also be derived (see Fig. S7). The slight difference in  could arise from the different synthetic ingredients, which result in disparate particle size and microstructures, etc.

Figure 3. Normalized SECM feedback approach curves for a 5 µm Pt disk electrode towards (a) LiFePO4 surface after different durations of chronoamperometric treatment. Experiment dots are fitted with irreversible substrate kinetics model. Extracted dimensionless rate constants (a) =

ACS Paragon Plus Environment

12

Page 13 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

1.46, 0.80, 0.45, 0.14. (b) Dependency of measured rate constant after CA pretreatment , with CA duration t, fit by error function; (c) illustration of LiFePO4/FePO4 transition inside the substrate during local transient mode experiment with left as before and right as after.

A further study was carried out to evaluate the relationship between the polarization time of the tip and the amount of the remained active material on surface as reflected by  . Feedback experiments were conducted on three separate spots of a 100% LiFePO4 substrate right after CA pretreatment for 60, 600 and 6000 s, respectively. As shown in Figure 3 (a), with increasing pretreatment time ! the positive feedback gradually attenuates due to the decreasing amount of active material remained after the tip polarization process. The non-linear relationship between the measured rate constants , after CA and pretreatment time ! is shown in Figure 3(b), which can be rationalized considering the reaction on the substrate have reached diffusionlimited region, such as phase boundary propagation or transport of charge carriers inside the materials. For the delithiation process, the CA pretreatment is expected to cause out-diffusion of Li+ from the sample at constant surface concentration and the concentration profile for one dimensional diffusion is given by the expression24 #

"#, = "$ − ("$ − "& )()*( ) √,

(5)

where "#, , Cs and C0 are the concentration of Li+ at detection depth -, surface of the sample and bulk concentration, respectively. D is the diffusion rate of Li+ inside bulk LiFePO4 matrix and ! is the pretreatment time. Under limiting diffusion condition, the relationship between the rate

ACS Paragon Plus Environment

13

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

constant , and that measured without CA treatment ,

.&

Page 14 of 19

can be derived as (see the

detailed analysis in Eq. S6-S9), /011,2

/011,234

=

56,2 54

#

= erf 8 √, 9

(6)

Fitting the experimental , to the above expression gives a good fit with R2=0.971 as shown in Figure 3(b), thereby demonstrating the validity of the above model. Furthermore, as the approach curve tip currents are proportional to the amount of active material, the CA pretreatment allows us to estimate the detection depth - of our SECM technique. Taking the Li+ diffusion coefficient of about 10-15 to 10-16 cm2/s,25 the detection depth during an approach curve is estimated to be about 1-3 nm of the sample surface. This agrees well with the lattice constants of the LiFePO4 unit cell and suggests that the probe technique is surface sensitive to a high degree. Another phenomenon worth noticing is that the positive feedback remains even after polarization for 600 s and 6000 s compared to substrate containing 0% active phase. The amount of electron transfer : can be estimated by integrating the tip current ; deducted by the theoretical pretreatment tip current at inert substrate ;