JP4728426B2 - Island-coated lithium cobaltite oxide - Google Patents

Description

The present invention relates to a powdered lithium transition metal oxide containing a special type of LiCoO ₂ supporting Mn and Ni. The positive electrode powder can be manufactured on a large scale by low cost processing. More specifically, the preparation is to sinter a cobalt-containing precursor, such as a mixture of LiCoO ₂ and Ni—Mn—Co-containing precursor, such as a mixed hydroxide MOOH and Li ₂ CO ₃ . The sintering temperature is high enough to allow cation exchange between LiCoO ₂ and the Li—Ni—Mn—Co oxide phase formed, and is very specific with a composition gradient of different transition metals. Form. The lithium transition metal oxide powder can be used in a rechargeable lithium battery as a positive electrode active material.

Despite some inherent limitations such as low safety and high cost, LiCoO ₂ remains the most applied positive electrode material for rechargeable lithium batteries. There is a strong demand driven by consumer expectations to increase the energy density of rechargeable lithium batteries. One way to improve energy density is to increase the charging voltage, which requires a more robust positive electrode material that can be charged at a higher voltage. The problems that appear or become more severe when increasing the charging voltage are: (a) low safety, (b) poor storage characteristics during storage of charged batteries at elevated temperatures, and (C) The cycle stability is poor. A number of approaches have been disclosed to address the problem. Partial improvements were achieved, but the basic problems were not fully resolved.

  In addition to the demand to increase energy density, it is essential that rechargeable batteries meet power requirements. This means that the battery as a whole has a sufficiently high rate performance, in particular that the active cathode material itself has a sufficiently high rate performance.

There is a general trend. A careful study of the published results for positive electrode materials can better understand the limitations of rechargeable lithium batteries based on LiCoO ₂ .

  One fundamental limitation stems from surface area issues. The improvement in rate performance (ie high power) can be accommodated by increasing the surface area. This is because the diffusion distance of lithium in the solid state can be reduced; resulting in improved rate performance. However, the large surface area increases the area where undesirable side reactions occur between the electrolyte and the charged cathode. Said side reactions are caused by the course of less safety, poor cycle stability at elevated voltage and poor storage properties of the charged cathode at elevated temperature. In addition, high surface area materials tend to have low recording densities that reduce the volumetric energy density.

Another basic limitation stems from the stoichiometric amount of cobalt. Cathode materials based on lithium-nickel-manganese-cobalt oxides (eg LiMn _1/3 Ni _1/3 Co _1/3 O ₂ ) are more stable to reactions between the electrolyte and the cathode than LiCoO ₂ And the raw material costs are lower, but the material suffers from a lower volumetric energy density and typically the material has a lower lithium diffusion constant.

It can be concluded that there are basic restrictions in the following respects:
Surface area: Cathode material with low surface area is desirable to achieve high safety, improved density and high stability during storage; however, the surface area is too low because rate performance will be reduced I can't.

Composition: LiMO ₂ positive electrode M, where M is primarily cobalt and is desirable to achieve high lithium diffusion rate and high volumetric energy density; however, high content of cobalt has been increased, less safety Leads to price and poor high voltage stability.

  One solution to this problem would be to increase the diffusion constant. Increased D could reduce the surface area without sacrificing rate performance.

LiMO ₂ , where M is Ni—Mn—Co, where Ni: Mn is greater than 1, as previously disclosed. For example, in US Pat. No. 6040090 (Sanyo), LiMO ₂ , where Ni: Mn is more than 1, LiMO ₂ containing (in this case, M is Mn, Ni, Co) ) A wide range is disclosed. This US patent specification discloses that LiMO ₂ has a high degree of crystallinity (HWFM with small peaks in the X-ray diffraction pattern). LiCoO ₂ doped with Ni and Mn is disclosed, for example, in US Pat. No. 7,078,128. This US Pat. No. 7,078,128 discloses that LiCoO ₂ doped with equal amounts of Ni and Mn is one preferred embodiment.

EP 1716609 A1 discloses an active cathode material based on LiMO ₂ , in which the composition of the particles depends on the particle size, in particular the cobalt content of the particles Decreases with decreasing diameter. The decrease in cobalt content originates from core-shell structured particles, where the Mn—Ni containing shell has the same thickness and coats the LiCoO ₂ core. As a result, when the particles are small, the LiCoO ₂ core is small and the cobalt content of all particles is low.

EP 1 556 915 A1 discloses LiMO ₂ with a transition metal composition gradient. This gradient is derived from the mixed hydroxide shell that coats the core with a significantly different metal composition. In a preferred embodiment, the core is LiCoO _2. After sintering, a transition metal composition gradient with a stoichiometric change in the radial direction is achieved, and the LiMO ₂ shell covers the core based on LiCoO ₂ . During sintering, cobalt diffuses from the LiCoO ₂ core into the LiMO ₂ shell. At the same time, Ni diffuses from the LiMO ₂ shell into the LiCoO ₂ core. Therefore, the shell expands and the LiCoO ₂ core contracts. An inflating shell that covers a shrinking core typically causes the formation of a void between the shell and the core. This void is highly undesirable.

The object of the present invention is to define a positive electrode material that has high rate performance and exhibits high stability during cycles extended at high charge voltages. High temperature storage properties are also improved. This is achieved by powdered lithium transition metal oxide containing LiCoO ₂ particles supporting Mn and Ni, in which case the particles have islands enriched in Mn and Ni content on the surface of the particles. And the island contains at least 5 mol%, preferably at least 10 mol% Mn.

The islands enriched in Mn and Ni content preferably have a thickness of at least 100 nm and cover less than 70%, preferably less than 50%, of the surface of the LiCoO ₂ particles supporting Mn and Ni. Further, Mn concentration in the islands, preferably LiCoO ₂ at least 4 mol% than the Mn concentration in the body of the particles supporting the Mn and Ni, preferably at least 7 mol% higher.

In another embodiment, the Ni concentration in the Mn and Ni enriched island is at least 2 mol%, preferably at least 6 mol, than the Ni concentration in the body of the LiCoO ₂ particles supporting Mn and Ni. %high. Preferably, the LiCoO ₂ particles supporting Mn and Ni contain at least 3 mol%, advantageously at least 10 mol% of both Ni and Mn. In one preferred embodiment, the crystal lattice constants a and c of LiCoO ₂ particles supporting Mn and Ni are 2.815 +/− 0.002 and 14.05 +/− 0.01, respectively.

In addition, LiCoO ₂ particles supporting Mn and Ni are monolithic and do not have internal porosity. Also preferably, the particle size distribution of the LiCoO ₂ particles supporting Mn and Ni has a d ₅₀ of _more than 10 μm, preferably more than 15 μm, most preferably more than 20 μm.

In another preferred embodiment, the powdered lithium transition metal oxide contains 30% to 95% by weight of LiCoO ₂ particles supporting Mn and Ni.

Further, the present invention provides a first phase composed of LiCoO ₂ particles supporting Mn and Ni, and further _a second island-free phase of the general formula Li _{1 + a} M ′ _1-a O _{2 ± b} . relates lithium transition metal oxide having, in this case, -0.03 <a <0.05 and b <0.02, M 'is _{Ni m Mn n Co 1-mn} , m> n and 0.1 <M + n <0.9. Furthermore, powdered lithium transition metal oxide, preferably has a total composition of _{Li x M y O 2 ± δ} , in this case 0.97 <x <1.03,0.97 <y < 1.03, x + y is 2, [delta] <0.05, M is _{Co 1-fg Ni f Mn g} , is 0.05 <f + g <0.5 and f ≧ g. It is also preferable that 0.98 <x / y <1.00. In another preferred embodiment, the oxide consists of only two phases, the first phase is LiCoO ₂ particles supporting Mn and Ni, and the second phase is an island-free phase.

The reference to island-free phase of the crystal lattice constant a 'and c' is constructed from an island-free phase corresponding to the same composition Li _x M _y O 2 has a _{± [delta]} and pure LiCoO ₂ particles The following relationship with the lattice constants a ″ and c ″ of the corresponding island-free phase of the lithium transition metal (M _ref ) oxide:
0.980 <a ′ / a ″ <0.998 and 0.9860 <c ′ / c ″ <0.9985, preferably 0.990 <a ′ / a ″ <0.997 and 0.9920 It is preferred to have <c ′ / c ″ <0.9980.

For example, if the material according to the invention, LiMO ₂ were prepared from mixed metal hydroxides of Co precursor and the composition formula _{M = Ni m Mn n Co 1} -mn , the lattice constant a '' and c '' is Due to the reference material having the composition LiM ″ O ₂ , the different lattice constants a ′ and c ′ are sufficient for the cations of the first phase based on LiCoO ₂ and the second phase free of islands. Clearly show that a new exchange has taken place.

The island-free phase preferably has a second particle having a particle size distribution with a d ₅₀ of 2 to 10 μm, which second particle has a particle size distribution with a d ₅₀ of 0.5 to 2 μm. It consists of sintered aggregates of primary crystallites.

Further, in another preferred embodiment, both the islands and the island-free phase the content of Mn and Ni-enriched, containing Ti, Ti content in this case, the oxide Li _x M _y O 2 _± Less than 10 mol% of M in _δ .

Further, more preferably, powdered lithium transition metal oxide, oxide Li _x M _y O 2 in _{± [delta],} wherein M of one or more dopants selected from the herd that is configured of Al and Mg And less than 1 mol% of said M of one or more dopants selected from the group consisting of Be, B, Ca, Zr, S, F and P.

For simplicity, LiCoO ₂ particles supporting Mn and Ni are referred to herein as “phase 1” or “modified LiCoO ₂ phase” in many cases of lithium transition metal oxides; The island-free phase having the general formula Li _{1 + a} M ′ _1-a O _{2 ± b} is the “LiM′O ₂ ′ phase” of the lithium transition metal oxide (where M ′ is Ni—Mn—Co). Or “phase 2” and this lithium transition metal oxide is also referred to as “positive electrode material”.

Surprisingly for the present invention, LiCoO ₂ (phase 1) and LiM′O ₂ ′ (where M ′ is Ni—Mn—Co), where the Ni: Mn ratio is above 1. The rate performance of the mixture with (Phase 2) is that the cation exchange occurs between LiCoO ₂ and LiM′O ₂ during the sintering of the mixture, and the composition distribution of the particles of Phase 1 and Phase 2 It has been disclosed that when treated together (co-sintered) to cause At the same time, a special form of phase 1 particles (LiCoO ₂ ) is obtained. The particles are partially covered by a manganese-containing LiM′O ₂ sheet. We refer to this form as the “island” form. At the same time, surprisingly, stability at high voltages is also significantly improved.

The modified LiCoO ₂ form has islands that are densely sintered against the modified LiCoO ₂ body, resulting in a local gradient of the stoichiometric amount of the transition metal. This island contains a high concentration of manganese. Both LiCoO ₂ and LiM′O ₂ particles have a single composition distribution. In addition, the LiM′O ₂ particles have a morphology that depends on the cobalt content. The size of the primary crystallite increases with the cobalt content. Unlike the above-mentioned EP 1 556 915 A1, there is no radial change in the stoichiometric amount in the present invention. Rather, it is a multi-centered gradient with LiM′O ₂ islands located on the surface that acts as the center of the gradient. Also, a partial coverage of LiCoO ₂ by islands is a very important difference.

Another important aspect of the present invention is that the island is not completely coated with LiCoO ₂ particles. A complete coating, in other words a LiCoO ₂ core-LiM′O ₂ shell morphology, can be achieved by precipitating the mixed hydroxide on the surface of LiCoO ₂ . This method is described in the above-mentioned European Patent Application Publication No. 1556915 A1 and European Patent Application Publication No. 1716609 A1 (Paulsen et al.). For the MOOH shell-LiCoO ₂ core precursor, Core-Shell Cathode Material with Size-Dependent Composition, Jens M. Paulsen, Jong-Seok Jeong, and Ki-Young Lee, Electrochem. Solid-State Lett., Volume 10, It has two main drawbacks as described in Issue 4, pp. A101-A105 (2007). (1) This method is even more expensive, and (2) during sintering, more cobalt diffuses from the core into the shell. Thus, the shell expands and the core contracts simultaneously. This typically causes partial separation of the shell from the core, resulting in large voids. This large void is highly undesirable. This is because (i) this void increases the porosity of the electrode and thus produces a much lower energy density, and (ii) this void into or into the core region of LiCoO ₂ particles This is because direct diffusion of lithium across the void from the core region is prevented, thus causing a loss of rate performance.

This situation is different from the positive electrode material according to the present invention. Manganese-containing islands cover only a portion of the surface of the LiCoO ₂ particles. Therefore, the diffusion of cobalt induced island expansion and LiCoO ₂ core contraction does not cause the formation of large voids. As a result, high volume density and high rate performance can be achieved.

  The present invention also relates to an electrochemical cell having a positive electrode containing the above-described powdery lithium transition metal oxide as an active material.

The method for producing the powdery lithium transition metal oxide described above includes the following steps:
LiCoO ₂ powder or a cobalt-containing precursor compound having a cobalt content of at least 90 mol% and Li-Ni-Mn-Co oxide or Ni-Mn-Co precursor powder and optionally a Li precursor compound, preferably lithium carbonate, Prepare a mixture of
Sintering the mixture at a temperature T of at least 900 ° C., preferably at least 950 ° C., for a time t of 1 to 48 hours,
It is thus possible to obtain LiCoO ₂ particles supporting Mn and Ni having islands enriched in the content of Mn and Ni on the surface.

Thus, the cathode material is a mixture of a LiCoO ₂ based powder and a Li—Ni—Mn—Co oxide or Ni—Mn—Co containing powder and a lithium source such as Li ₂ CO ₃ that is higher than 900 ° C. Manufactured by sintering at temperature. This temperature must exceed 900 ° C., for example 910 ° C. or 920 ° C. During sintering, a partial exchange of cations takes place between the LiCoO ₂ particles and the Ni—Mn containing particles. When the sintering temperature is low, the cations are not sufficiently exchanged, and the positive electrode does not exhibit high rate performance. When the sintering temperature is high, the particles are too dense, the metal composition becomes excessively equilibrium, i.e. cation exchange between the LiCoO ₂ and Mn-Ni-Co is excessively performed. In this case, there will be no islands enriched with Mn and Ni content on the first phase particles.

  In the case of a selectable method, a cobalt-containing precursor powder (e.g., cobalt oxide, cobalt hydroxide or cobalt carbonate) can be mixed with the Ni-Mn-Co-containing powder and a lithium source, and continue to high temperatures. , Preferably at temperatures above 950 ° C.

The method for producing the above-described powdery lithium transition metal oxide having two phases includes the following steps:
LiCoO ₂ powder or a cobalt-containing precursor compound having a cobalt content of at least 90 mol% and Li-Ni-Mn-Co oxide or Ni-Mn-Co precursor powder and optionally a Li precursor compound, preferably lithium carbonate, Prepare a mixture of
Sintering the mixture at a temperature T of at least 900 ° C., preferably at least 950 ° C., for a time t of 1 to 48 hours,
Thus, an LiCoO ₂ particle phase supporting Mn and Ni and an island-free phase with crystal lattice constants a ′ and c ′ are obtained, in which case these phases are Li—Ni—Mn—Co oxide, or Ni— With the lattice constants a ″ and c ″ of the reference lithium transition metal (M _ref ) oxide obtained by sintering the Mn—Co precursor powder and the Li precursor compound at the same temperature T and the same time t Having the following relationship:
In this case, this relationship is
0.980 <a ′ / a ″ <0.998 and 0.9860 <c ′ / c ″ <0.9985, preferably 0.990 <a ′ / a ″ <0.997 and 0.9920 <C ′ / c ″ <0.9980.

In this method, the Ni-Mn-Co precursor powder is preferably a transition metal hydroxide, transition metal oxyhydroxide, transition metal carbonate, transition metal oxycarbonate, or lithium transition metal compound, In this case, the transition metal composition M ″ is M ″ = Ni _o Mn _p Co _1-op , where o + p> 0.5 and o> p. The Ni-Mn-Co precursor powder preferably contains a transition metal content of 5 to 70 mol% of the powdered lithium transition metal oxide.

In one embodiment, the LiCoO ₂ powder used is composed of monolithic particles having a tap density of at least 2 g / cm ³ and having a d ₅₀ of at least 10 μm, preferably at least 15 μm, most preferably at least 20 μm. ing.

  On the other hand, the cobalt-containing precursor compound is preferably one of cobalt hydroxide, cobalt oxyhydroxide or cobalt carbonate.

In another embodiment, the LiCoO ₂ or cobalt-containing precursor contains at least 80% of the transition metal of the powdered lithium transition metal oxide and the Ni—Mn—Co-containing precursor powder has a d ₅₀ of 1-3 μm. It is comprised from the particle | grains which have the particle size distribution which has.

In another embodiment, the LiCoO ₂ or cobalt containing precursor contains less than 80% of the transition metal of the powdered lithium transition metal oxide and the Ni—Mn—Co containing precursor has a d ₅₀ of 4-10 μm. It is composed of aggregated type particles having a particle size distribution.

In both of these embodiments, the Ni—Mn—Co containing precursor can further comprise Ti, preferably in the form of TiO ₂ particles having a d ₅₀ of less than 100 nm.

  Further details of the invention are described below.

The positive electrode material of the present invention is a powder containing modified LiCoO ₂ and, in many cases but not exclusively, a second transition metal phase. Both phases are lithium transition metal oxide phases having a layered crystal structure. Regular rock-salt crystal structure Space group r-3m. The positive electrode can be a stoichiometric amount of Li ₁ M ₁ O ₂ , where M is cobalt, manganese and / or nickel, or is slightly deficient in lithium (Li _{1− x} M _{1 + x} O ₂ ) or lithium rich Li _{1 + x} M _1-x O ₂ , where x <0.3. The presence of non-stoichiometric amounts of oxygen is generally questionable. That is, the stoichiometric amount of oxygen is about 2.0, but it cannot be excluded that the positive electrode is slightly oxygen deficient or rich in oxygen. Therefore, the total composition is _{Li x M y O 2 ± δ} , is in this case 0.97 <x <1.03,0.97 <y < 1.03, x + y = 2 and [delta] <0.05 . M is manganese, which is composed of cobalt and _{nickel, M = Co 1-fg Ni} f Mn g, however, assumed in this case is 0.05 <f + g <0.5 and f ≧ g.

The first phase is derived from LiCoO ₂ precursor and a modified LiCoO _2. This composition can be defined as LiCo _1-ab Ni _a Mn _b O ₂ , where a ≧ b, 0.03 <a + b <0.5, preferably 0.1 <a + b <0.5 Shall. The formula is considered ideal, but small expected deviations are not taken into account, such as excess or deficiency of lithium, non-stoichiometric amount of oxygen or doping as described above. Preferred LiCoO ₂ based particles are monolithic. Monolithic particles do not exhibit internal porosity and are not composed of even smaller primary particle aggregates. One aspect of the present invention is that the different particles of the LiCoO ₂ phase do not have exactly the same composition. The actual composition of the particles depends on how much nickel and manganese are diffused into LiCoO ₂ during sintering. Ni and Mn are derived from a second phase precursor, which is typically a mixed hydroxide. The amount of Mn and Ni that diffuses into the LiCoO ₂ based phase during sintering, along with a number of other factors such as temperature, Li: M ratio, etc., the placement of adjacent Ni—Mn based particles And depends on the contact area and the contact pressure. As a result, different LiCoO ₂ particles have different compositions.

A second very important aspect of the present invention is that the metal composition of one LiCoO ₂ based particle is non-uniform. Typical particles have an island like surface morphology, the island this case was densely sintered on the surface of the LiCoO ₂ particles, particles or crystallites which are based on more small Ni-Mn Derived from. This island has a higher concentration of manganese than the region further away from the island or the region inside the particle. The presence of the island morphology is one unique feature of the positive electrode material of the present invention. The island is a central part with an even higher manganese content and cannot be separated from the particles. The islands are densely and continuously bonded to the body of LiCoO ₂ particles. As the distance from the islands increases, the stoichiometric amount of manganese approaches in the interior of the particle, possibly in a gradient fashion, or approaches zero on the surface between spaced islands. The inventors have observed that the island morphology is related to the high observed rate performance of the disclosed positive electrode material. We speculate that islands will have different crystal lattice constants if they are not bound to LiCoO ₂ particles. However, islands are densely coupled to LiCoO _2, between the LiCoO ₂ particles and the island, there is a region of the gradient of the stoichiometric amount of manganese. Therefore, the islands as well as the particles will experience strong lattice distortions. This strain allows the remarkably fast diffusion of lithium into the particles, although the exact mechanism is not known to the inventors.

The second phase is Li ₁ M'O _2, in this case M 'is _{Ni m Mn n Co 1-mn} , m ≧ n, it is assumed that 0.1 <m + n <0.9 . The formula is considered ideal, but small expected deviations are not taken into account, such as excess or deficiency of lithium, non-stoichiometric amount of oxygen or doping as described above. The second phase is preferably derived from a Ni-Mn-Co containing precursor, such as a mixed hydroxide, mixed oxyhydroxide, mixed oxide, mixed lithium metal oxide or mixed carbonate. During sintering, the metal composition of the second phase changes. Cobalt diffuses from LiCoO ₂ particles into Li ₁ M′O ₂ particles. Some Ni and Mn diffuse from Li ₁ M′O ₂ particles into LiCoO ₂ particles. As a result, the stoichiometric amount of cobalt in the second phase is higher than the stoichiometric amount of cobalt in the Ni-Mn-Co containing precursor. The change in the stoichiometric amount of cobalt is one important aspect of the present invention. Only when the stoichiometric amount of cobalt increases sufficiently during sintering, sufficient exchange of cations occurs, and the rate performance of the positive electrode that occurs only in this case is sufficiently improved.

  The inventors have observed two more surprising facts that are considered to be a more essential point of view of the present invention.

First observation: The fraction of the second phase increases during sintering. Apparently, cobalt diffuses in the _second phase (Li ₁ M′O ₂ ) rather than nickel, and manganese diffuses in the LiCoO ₂ phase. We speculate that this difference in diffusion increases the observed island morphology. Associated with the observation is a clear change in voltage profile. The mixture of LiCoO ₂ and Li ₁ M′O ₂ has a characteristic voltage profile with a plateau at 3.88V. As cations increased, we observed the disappearance of the 3.88 V plateau with decreasing final discharge voltage. Furthermore, cobalt not only diffuses into the Li ₁ M′O ₂ particles, but also diffuses into the manganese-containing region on the surface; during this process, the region between the islands acts as a Co source. At the same time, the island itself is a cobalt sink. In one simple image, a manganese-containing island expands with cobalt, for example, as a sponge expands by removing water from the periphery of the sponge. This method explains why the island morphology is formed.

Second observation: The first phase has a composition that is clearly different from pure LiCoO ₂ . The majority of the first phase particles contain at least 3%, preferably 10%, of manganese and nickel. Such a stoichiometric change is usually achieved by significant changes in the lattice constant. However, the X-ray diffraction analysis surprisingly shows that the lattice constant of the first phase (obtained by the improvement of the two-phase Rietveld) is essentially unchanged, that is, the lattice constant is the same as that of LiCoO2. Indicates that it remains. We are in a very important aspect of the invention that this shows that the improvement in rate performance of the first phase is not caused by the formation of a solid solution between LiCoO ₂ and Li ₁ M′O _2. I think there is. (The solid solution shows a stepwise change in lattice constant depending on the composition.) Another aspect of the present invention is that the Li ₁ M′O ₂ particles (second phase) have crystallites, and crystallites. The size of the rit is correlated with the cobalt content. Obviously, during sintering in which more Ni (and Mn) diffuses from Li ₁ M′O ₂ particles into LiCoO ₂ particles and more Co diffuses into Li ₁ M′O ₂ particles, Crystallite growth is accelerated. As a result, Li ₁ M′O ₂ particles (second phase) with higher cobalt stoichiometric amounts have a much larger primary crystallite. This is a very useful method. This is because an optimized form is achieved in a self-managed manner. This allows for even greater crystallites without loss of rate performance since the increased content of cobalt causes faster lithium diffusion. However, the correlation between high cobalt content and larger dimensions is only attributed to crystallite dimensions, not particle size. Larger particles appear to have an average lower cobalt stoichiometry than smaller particles. This is because more cobalt has to diffuse longer paths.

We understand the reaction that causes island morphology as follows:
During sintering, a significant fraction of smaller Li ₁ M′O ₂ particles and agglomerated Li ₁ M′O ₂ particles are contacted with LiCoO ₂ particles. The contact point is a cobalt sink, which forms a manganese-containing island that is inherently embedded on the surface of the LiCoO ₂ particles. At the same time, nickel (and some manganese) diffuses into the LiCoO ₂ particles and cobalt diffuses into the Li ₁ M′O ₂ particles. During sintering, the density of agglomerated Li ₁ M′O ₂ particles, caused by the absorption of cobalt, increases due to thermal sintering. During densification, contact between the expanding islands and the Li ₁ M′O ₂ particles is lost and a final positive electrode composed of two different phase particles is achieved.

Loss of contact between Li ₁ M′O ₂ and LiCoO ₂ is simplified when Li ₁ M′O ₂ particles are aggregated. In this case, only a portion of the Li ₁ M′O ₂ particles are consumed, forming a seed for the island. According to another selectable manner, Ni-Mn-Co precursor when having very small particles having a d ₅₀ of less than 1~2μm is no loss of contact is required. In this case, most of the Ni—Mn—Co particles are consumed, or even the entire Ni—Mn—Co particles are consumed to form island seeds. As a result, different embodiments of the invention are possible.

First exemplary embodiment: It is particularly preferred that the Ni—Mn—Co precursor is composed of aggregated crystallites. One preferred example is a mixed hydroxide, in which case the secondary particles are composed of aggregates of non-fine primary particles. Very dense and large Ni-Mn-Co precursors are not very suitable. A preferred particle size distribution has a d ₅₀ of 4-8 μm. In this case, the Li ₁ M′O ₂ particles are (a) small enough to support very high velocities, and (b) the particles allow for low porosity electrodes and high volumetric energy density. It fits nicely into the gaps of even larger particles.

Preferably, the precursor for the first phase (LiCoO ₂ ) is monolithic and dense, for the second phase (Li ₁ M′O ₂ ) that is agglomerated and less dense but of smaller dimensions. Have significantly larger dimensions than the precursors. A preferred precursor for the first phase is LiCoO ₂ with dense monolithic particles of at least 10-20 μm. Many commercial LiCoO ₂ materials have this desirable form. According to other alternative methods, cobalt hydroxide, cobalt oxyhydroxide, cobalt oxide or cobalt carbonate are suitable precursors if they have large particles (at least 10-20 μm) and high density. As one example, cobalt hydroxide or cobalt oxyhydroxide having rough spherical particles and a tap density of greater than 2.0 g / cm ³ and a d _{50 of} particle size distribution greater than 15-20 μm are suitable precursors. .

Ni-Mn-Co precursor is agglomerated, and if it has a particle size distribution with a d ₅₀ of 4~10μm is preferably at least 20% of the transition metal in the final positive electrode, Ni-Mn-Co precursor And less than 80% of the transition metal is derived from the LiCoO ₂ precursor.

Second exemplary embodiment: It is also preferred that the Ni-Mn-Co precursor is composed of very small particles. One example is a jet milled mixed hydroxide with typical particles below 0.5-1.5 μm. In this case, advantageously less than 20% or even 15% of the final positive electrode transition metal is derived from the Ni—Mn—Co precursor, whereas at least 80%, preferably 85%, Derived from a cobalt precursor. The cobalt precursor is composed of large particles (d ₅₀ > 10 to 20 μm) which are advantageously dense and monolithic. Suitable cobalt precursors are commercial LiCoO ₂ or high density (tap density> 2 g / cm ³ ) cobalt hydroxide, cobalt oxyhydroxide or cobalt carbonate. Suitable shapes for the precursor are, for example, spherical or irregular potato-shaped particles.

The two typical implementations are not considered other selectable examples, but rather are considered two extreme examples. For example, using a Ni—Mn—Co precursor having a bimodal particle size distribution and containing small agglomerated particles (less than 0.5-1.5 μm) and large agglomerated particles (4-8 μm) It is possible, in this case the majority of the small particles are consumed and form islands, the majority of the large particles being cut during sintering. It is also possible to use even smaller cobalt particles and submicrometer MOOH, in which case very high rate performance can be expected. The formation of the manganese-containing islands achieved by the reaction, ie the cation exchange between cobalt and nickel, is the same in both embodiments described above. We believe that the essential point of view causing the formation of island morphology is the lower mobility of (tetravalent) manganese compared to the mobility of trivalent nickel and trivalent cobalt in LiCoO _2. thinking. Also, (tetravalent) manganese contributes to electrochemical insertion / extraction of lithium during battery charging / discharging. Some manganese may be replaced by other cations. A suitable cation is titanium. This titanium, like manganese, is electrochemically inert, has low mobility, and may be doped into Ni-Mn-Co precursors. For example, like manganese, titanium may be doped in LiNiO ₂ .

Another important aspect of the present invention is that high rate performance is achieved even though the cathode material is slightly lithium stoichiometrically deficient. We have observed that the highest rate performance is achieved even though the total lithium content per transition metal is about 0.98, or less than one unit. This is extremely phenomenal. This is because the lithium transition metal oxide Li _{1 + z} M _1-z O ₂ , where M contains nickel, and in this case, the lack of lithium causes cation mixing (this is due to the presence of crystalline lithium sites). This is because it is widely accepted that increased cation mixing causes poor rate performance.

  The figures illustrating the present invention are summarized as follows:

SEM micrographs of samples REF1 and REF2. The SEM micrograph of mixed sample CX2. SEM micrograph of sample CX3. SEM micrograph of sample EX1. SEM micrograph sample of sample EX3. SEM micrograph of phase 1 of sample EX1. SEM micrograph of sample EX2-phase 1 and 2. SEM micrograph of EX1 particles for EDS analysis. EDS mapping for phase 1 particles of EX1. SEM micrograph of EX1 particles for EDS analysis. SEM micrograph of EX1 particles for EDS analysis. Cyclic behavior of commercially available LiCoO ₂ (REF1) and sample EX4. SEM micrograph of sample EX5E. SEM micrograph of sample CX6. Crystal map of REF1-2, CX2-3 and EX1-3. Crystal map of REF1-2, CX5 and EX4-5. X-ray diffraction patterns of CX2, CX4 and CX5, and EX1. X-ray diffraction pattern of CX6 and EX9E. Voltage profiles of CX2, CX3 and EX1-EX3 during slow discharge. Cycle behavior and rate performance of sample EX1. The rate performance of sample CX6 is compared to EX5E.

  In the following examples, some aspects of the present invention will be further described. The following table summarizes the test conditions and test results.

  Table 1 provides a summary of the samples and manufacturing conditions. Table 2 contains X-ray and BET surface area data.

  Table 3 gives an overview of the electrochemical results obtained from the coin-type battery.

Reference Example The following reference sample was used:
REF1 LiCoO ₂ is commercially available LiCoO ₂ , has a d ₅₀ of about 20 μm, and is composed of monolithic dense particles.

REF2 LiM′O ₂ is produced from mixed hydroxide MOOH and Li ₂ CO _{3 in} air at 950 ° C .; the Li: M ratio is Li: M ′ = 1.10: 1 and M ′ = Ni _0.53 Mn _0.27 Co _0.2 . REF2 has an aggregated form.

  Both samples REF1 and REF2 were reheated at 850 ° C. for 8 hours before coin cell assembly and BET measurement. Measure X-ray diffraction pattern and improve Rietveld. FIG. 1 shows SEM micrographs of samples REF1 and REF2. The figure on the left shows REF1 at 1000 magnification. The particles have an irregular shape. There is no island form. The figure on the right shows REF2 at 2500 magnification. The particles are composed of primary crystallites that are agglomerated and sintered into secondary particles having large irregular shapes.

Calculation Example For a hypothetical calculated sample CC1, which is a mixture of 60% REF1-LiCoO ₂ and 40% REF2-Li ₁ M′O ₂ , the expected values for BET surface area, discharge capacity and rate performance are , By evaluating the mass average of the corresponding values of REF1 and REF2.

Comparative Example Example CX2: A positive electrode powder is prepared by mixing REF1-LiCoO ₂ 60% and REF2-Li ₁ M′O ₂ . Prior to mixing, both REF1-LiCoO ₂ and REF2-Li ₁ M′O ₂ were heated in air at 850 ° C. for 5 hours. The total composition of the final CX2 positive electrode is Li ₁ M′O ₂ where M ′ is Co _0.68 Ni _0.21 Mn _0.11 . FIG. 2a shows a SEM micrograph (5000 magnifications) of the mixed sample CX2. The BET surface area of the mixed powder CX2 is measured. The island morphology cannot be observed. A coin-type battery is manufactured and measured for discharge capacity, irregular discharge capacity, cycle stability and rate performance. Measure X-ray diffraction pattern and improve Rietveld. Take a SEM micrograph.

Tables 2 and 3 show that sample CX2 has approximately the same properties as the mass average of the precursor in hypothetical sample CC1. This mixing does not provide a significant advantage in rate performance or cycle stability. SEM micrographs confirm the absence of island morphology of LiCoO ₂ particles. Rietveld's improvement confirms that the lattice constant obtained from the X-ray diffraction pattern of the mixture is the same as that obtained from the X-ray diffraction pattern of LiCoO ₂ and Li ₁ M′O ₂ , respectively.

Example CX3: A positive electrode powder is prepared by mixing REF1-LiCoO ₂ 60% and REF2-Li ₁ M′O ₂ 40%. This mixture is heated in air at 850 ° C. for 5 hours, yielding sample CX3. The total composition of this positive electrode is Li ₁ M′O ₂ where M ′ is Co _0.68 Ni _0.21 Mn _0.11 and is the same as CX2. FIG. 2b shows a SEM micrograph of sample CX3. The magnification is 2500 times. There is no island form.

Properties such as cycle stability and rate performance of sample CX3 (which is a heat-treated mixture) are clearly slightly improved compared to CX2 (which is a heat-treated sample). The Rietveld improvement confirms that the lattice constants of the compounds LiM′O ₂ and LiCoO ₂ were not significantly changed during the heat treatment. The constant of REF1 is the same as the constant of phase 1 in CX2 and CX3, and the lattice constant of REF2 is the same as the lattice constant of phase 2 of CX2 and CX3.

  Example CX4: A positive electrode powder, a heat-treated mixture, is prepared similarly to the method described in Comparative Example CX3, but is heated at 900 ° C. for 5 hours instead of 850 ° C. for 5 hours, yielding sample CX4 . A coin-type battery is manufactured. Measure X-ray diffraction pattern and improve Rietveld. Take a SEM micrograph.

Tables 2 and 3 show that sample CX4 has approximately the same properties as CX3 produced at low temperatures. SEM micrographs show that the island morphology is essentially absent. X-ray diffraction shows a two-phase phase mixture, where the first phase has a lattice constant of REF1-LiCoO ₂ and the second phase is the same lattice as the sample REF2-Li ₁ M′O _2. Have a constant. Clearly no effective diffusion of Co from phase 1 LiCoO ₂ to the second phase Li ₁ M′O ₂ occurred. The rate performance is the same as that of the sample CX3. This comparative example shows that increasing the heat treatment temperature from 850 ° C. to 900 ° C. does not provide any significant improvement in the performance of the coin cell battery.

Examples of the Invention Example 1 (EX1): Cathode powder is prepared by mixing 60% commercial LiCoO ₂ (sample REF1) with mixed transition metal hydroxides MOOH and 40% Li ₂ CO ₃ . The Li ₂ CO ₃ : MOOH ratio and the mixed hydroxide are the same as those used for the production of REF2-Li ₁ M′O ₂ . The total composition of this positive electrode powder is Li ₁ M′O ₂ in which M ′ is Co _0.68 Ni _0.21 Mn _0.11 , and is the same as the total composition of CX2 and CX3. This mixture is heated in air at 970 ° C. for 8 hours, yielding sample EX1.

A coin-type battery is manufactured. Measure X-ray diffraction pattern and improve Rietveld. Take a SEM micrograph. FIG. 3a shows a SEM micrograph of sample EX1. The magnification is 5000 times. There are two types of particles: (a) Phase 1: particles based on LiCoO ₂ with a dense and irregular shape, specially with island morphology and (b) Phase 2: Agglomerated type Li ₁ M′O ₂ particles: The primary crystallite size has an expanded distribution. Phase 1 is clearly shown in FIG. 3c. EDS analysis (see below) highlights the presence of Mn in the islands on the surface of the modified LiCoO ₂ particles.

  Properties such as cycle stability and rate performance are significantly better than hypothetical sample CC1 and are significantly improved when compared to samples CX2 and CX3.

SEM micrographs confirm the presence of island morphology of LiCoO ₂ particles. It is confirmed that, due to the improvement of Rietveld, the lattice constant of phase 1 (LiCoO ₂ ) does not change during the heat treatment, but the lattice constant of phase 2 (Li ₁ M′O ₂ ) changes significantly. The change in the lattice constant of Li ₁ M′O ₂ proves that a significant exchange of cations between phase 1 and phase 2 was associated.

  EX2 and EX3 Examples: Cathode powders were manufactured and studied in the same manner as EX1 in Example 1, but the sintering temperatures were 960 ° C. and 950 ° C., respectively (sintering time: 8 hours). FIG. 3b shows a SEM micrograph of sample EX3. FIG. 4 shows SEM micrographs of the two phases of sample EX2: the diagram on the left shows mainly Phase 2 particles, the diagram on the right shows mainly Phase 1 particles, in this case, It can also be seen that the Phase 1 particles are significantly larger than the even smaller Phase 2 aggregates.

  Also, properties such as cycle stability and rate performance are significantly better than the hypothetical sample CC1, and are significantly improved when compared to samples CX2 and CX3.

SEM micrographs confirm the presence of island morphology of LiCoO ₂ particles. With Rietveld improvement, it is confirmed that the lattice constant of LiCoO ₂ does not change during the heat treatment, but the lattice constant of Li ₁ M′O ₂ has changed significantly. Compared to Examples 1, 2, and 3, this change can include that it is more important at higher temperatures, in which case (a) in Li ₁ M′O ₂ It is shown that the amount of Co diffusion into the layer increases with temperature, but at the same time (b) the improved properties do not sensitively depend on the amount of Co in the Li ₁ M′O ₂ phase.

EDS Analysis of Samples When using energy dispersive X-ray spectroscopy (EDS), the composition of samples CX2 and CX3 (comparative examples) and Example EX1 can be studied.

  EDS analysis is a powerful method for studying the composition of particles near the surface. EDS is particularly powerful for monitoring changes and trends, but not very powerful for obtaining accurate quantitative results. Table 4 discloses the results of EDS analysis of reference samples REF1 and REF2, which will be used as reference examples for EDS analysis of more complex samples CX2, CX3 and EX1.

Sample REF1 (LiCoO ₂ ) was studied by EDS spectroscopy. Spectra measured from a number of particles were collected. The magnification is 1000 times, and the scanned area is the area shown in FIG. Similarly, the EDS spectrum of sample REF2 was collected at a magnification of 1000 times.

Comparison of the results obtained from ICP chemical analysis and EDS analysis shows that EDS (1) estimates the transition metal ratio almost accurately,
(2) Indicates that the sulfur content is emphasized (sulfur impurities that appear to be located on the surface).

  The positive electrode sample EX1 was studied by applying EDS analysis to a single particle. EDS spectra of 6 different particles of phase 1 were obtained. All particles showed island morphology. A SEM micrograph of the six particles is shown in FIG.

EDS analysis clearly shows that phase 1 (LiCoO ₂ ) particles contain large amounts (> 15%) of nickel and manganese (see Table 5 below). This is extremely phenomenal. This is because the Rietveld improvement of the X-ray diffraction pattern showed that phase 1 (containing NI and Mn) has the same lattice constant as LiCoO ₂ . Furthermore, 5 of the 6 particles have a Ni: Mn ratio greater than 3.0. This indicates that more nickel than manganese has diffused into the first phase. During the sintering, cation exchange was performed. In this case, nickel mainly entered from the Li ₁ M′O ₂ particles into the LiCoO ₂ particles together with manganese. EDA analysis confirms that the first phase (LiCoO ₂ ) particles have a composition distribution with a transition metal composition that varies.

  Two of the six particles in Table 5 (particle No. 1 and particle No. 2) were studied by EDS mapping. In FIG. An EDS mapping of 1 indicates that the “islands” have a high content of manganese, whereas the inter-island region, “oceans” (or body) has a low content of manganese. Furthermore, the particle No. 4 and no. 6 was studied by spot EDS analysis (see Table 6). FIG. 7 shows the position of the spot. The spectrum of the spot was collected.

  All “island” spots (X2, X4, X6, X7) have a clearly lower Ni: Mn ratio than for all particles (Table 5). All “ocean” spots (X5, X8, X9) have a significantly lower manganese content than the whole particles. The examples confirm that particles with island morphology have a high Mn content in most islands and a low manganese content between islands. Clearly there is a manganese gradient with an island that is the center of the gradient.

The EDS spectra of three single particles of the second phase (Li ₁ M′O ₂ ) of sample EX1 were collected. The particles are derived from MOOH having the same metal composition as sample REF2, in which case the Ni: Mn ratio is about 2.0 and the cobalt content is about 20%. FIG. 5 shows a SEM micrograph. The three particles have crystallites of clearly different dimensions. Particle 1 (left side) has a crystallite of about 0.5-1.5 μm; particle 2 (center) has a crystallite of about 1-2 μm, and particle 3 (right side) has a crystallite of about 1. It has a crystallite of 5 to 3 μm. Similarly, EDS spectra of single Li ₁ M′O ₂ particles (phase 2) of samples CX2 and CX3 were collected. All results are reported in Table 7.

The cobalt content of the second phase Li ₁ M′O ₂ particles of sample EX1 increased significantly during sintering. This is in sharp contrast with the results for the Li ₁ M′O ₂ particles of Samples CX2 and CX3, which have approximately the same EDS spectrum as Sample REF2. This observation indicates that cation exchange takes place during the sintering of EX1, where cobalt from LiCoO ₂ (phase 1) has penetrated into the Li ₁ M′O ₂ (phase 2) particles. Furthermore, a comparison between the SEM micrograph in FIG. 8 and the data in the table shows that the primary crystallite size of sample EX1 and the cobalt content of phase 2 particles are correlated. Obviously, if the cobalt diffuses into the Li ₁ M'O ₂ is sinterability of Li ₁ M'O ₂ particles are strengthened, causing Yorisso rapid crystallite growth.

Example 4: Jet Mill Milling of Precursors A mixed hydroxide of submicrometer size was produced by jet milling a mixed hydroxide MOOH. MOOH is the same as that used for the production of REF2-Li ₁ M′O ₂ . The particle size distribution was measured by laser diffraction. After three jet milling operations, 80% of the volume is composed of particles having a dimension of less than 1 μm.

90% by weight of commercially available LiCoO ₂ particles (sample REF1, with 20 μm particles) and 3% jet milled MOOH 10% were mixed with Li ₂ CO ₃ . 1/2 mole of Li ₂ CO ₃ was added to 1 mole of MOOH pulverized by jet mill. (The Li: M ratio is the same as the ratio used for the production of REF2-Li ₁ M′O _2. ) After mixing, the sample was sintered at 970 ° C. for 8 hours.

  The final sample EX4 was studied by SEM, BET surface area analysis and X-ray diffraction. A coin-type battery was manufactured. Rate performance and cycle stability were measured.

FIG. 9 compares the rate performance (battery voltage V versus discharge capacity mAh / g) of the commercially available LiCoO ₂ (REF1) on the left side (A) with the rate performance of the sample EX4 on the right side (B). This figure shows the discharge voltage profile during C / 10, C / 5, C / 2, 1C. The 1.5C, 2C, 3C, 5C and 10C rates, in this case 1C (corresponding to 1 hour discharge), are defined as 160 mAh / g. The temperature was kept constant at 24 ° C. and the voltage range was 4.3-3.0V. Clearly, the rate performance has increased dramatically. SEM micrographs (not shown) clearly show the presence of island morphology.

Study of co-sintering conditions Sample CX5 was produced in the same manner as samples EX1, EX2 and EX3, but the sintering temperature was reduced to 900 ° C. (sintering time: 8 hours). The sample was clearly different from EX1, EX2, EX3. The BET surface area was large enough: 0.35 m ² / g. X-ray diffraction shows a phase mixture of the two phases, the first phase has the lattice constant of REF2-Li ₁ M′O ₂ and the second phase is the sample REF2-Li ₁ M′O _2. Have the same lattice constant. Clearly, no effective diffusion of Co from phase 1 LiCoO ₂ to the second phase Li ₁ M′O ₂ occurred. Similarly, the volume fraction of the second phase is clearly negligible, in this case that this volume fraction is almost free of Co diffused into phase 2 (Li ₁ M′O ₂ ). I'm doing it.

The electrochemical properties are poor (Table 3). Poor cycle stability is observed (fading rate at 4.5V is almost 2-3 times faster than for samples EX1-EX3). The rate performance is significantly lower (87.5% at 3C rate) compared to 90-91% for samples EX1, EX2, EX3. The speed performance is the same as that of the sample CX3. SEM micrographs (not shown) show slightly smaller Li ₁ M′O ₂ particles bound on the surface of even larger LiCoO ₂ , but the island morphology is essentially absent.

The positive electrode powder is manufactured and analyzed in the same manner as the positive electrode powder of Example 4. However, different precursors were used for the second phase Li ₁ M′O ₂ . In this example, REF1-LiCoO ₂ 90% is mixed with jet milled precursor and 0.05 mol% Li ₂ CO ₃ 10%. The precursor is a lithium stoichiometric deficiency of Li _1-X M _{1 + X} O ₂ . The precursor was produced in the same manner as REF2-Li ₁ M′O ₂ but the Li: M ratio was 0.9 and the temperature was 900 ° C. After manufacture, the precursor was jet milled twice to produce a submicrometer granular product. The particle size distribution was measured by laser diffraction in water. The particle size distribution is bimodal, where about 50% of the volume has a dimension of 0.05-1 μm (maximum at about 0.3 μm) and the remaining 50% of the volume is 1-6 μm. (Maximum at about 2 μm). The mixture was heated in air at 970 ° C. for 8 hours. Measure X-ray diffraction pattern and improve Rietveld. Take a SEM micrograph. A coin-type battery is manufactured.

The X-ray diffraction pattern basically shows one phase having the same lattice constant as LiCoO ₂ . The second Li ₁ M′O ₂ is clearly indistinguishable. (This is different from the sample of Example 4, which clearly shows the presence of the second phase). FIG. 10b shows a SEM micrograph. There are very few Li ₁ M′O ₂ particles (= phase 2) of the agglomerated type. Almost all particles are based on LiCoO ₂ (= phase 1). The particles generally have a very smooth surface. Obviously, the island form is absent. The observed very low BET surface area, which is only 0.14 m ² / g, is consistent with the above observations.

Clearly, sample CX6 was sintered more efficiently than sample EX4. Probably too much cobalt diffused from phase 1 LiCoO ₂ to phase 2 Li ₁ M′O ₂ . At the same time, a small number of Li ₁ M′O ₂ particles are consumed by the large LiCoO ₂ particles, presumably the manganese cations in LiCoO ₂ are diluted and, as a result, the island morphology is absent. The compositions of Phase 2 and Phase 1 were effectively close to each other. Even though the second phase has a much larger fraction of the positive electrode than in the almost unsintered sample, it is now very similar to phase 1 and these phases are, for example, X It can no longer be clearly distinguished by the line.

Electrochemical tests show that:
(A) The slope of the voltage profile at the end of the discharge disappeared. This is consistent with the essentially absence of phase 2 Li ₁ M′O ₂ .

  (B) Rate performance is significantly lower than that of sample EX4, and (c) cycle stability is poor.

It can be concluded that the presence of the island morphology and the second phase is essential for obtaining high rate performance. Furthermore, there is a fairly narrow window to achieve a high rate positive electrode. If the sintering is too strong (sample CX6), the islands disappear. This is because when the sintering is not sufficient (samples CX3 and CX4), the diffusion amount of the transition metal is high and the diffusion of the transition metal is insufficient, so that no island is formed. Tables 2 and 3 summarize the data obtained. Indeed, in order to carry out the method according to the invention, it is necessary to establish the temperature versus the sintering temperature, in which case the SEM micrographs of the product obtained show an island structure of EX1-4. If no co-sintering occurred, phase 2 is clearly distinguished and pure LiCoO ₂ is observed without island morphology. If the co-sintering occurs too strongly, phase 2 is almost lost and the resulting Li-Co-Ni-Mn oxide has a smooth surface with circular edges.

If two phases LiCoO ₂ and Li ₁ M′O ₂ are present, it is necessary to measure the lattice constant of the resulting sample and to obtain phase 2 (the absence of LiCoO ₂ or the corresponding cobalt precursor) It is possible to compare with a reference sample, which is a sintered compound obtained only with the precursor of The relationship between the obtained lattice constants is within the limits cited above.

Effect of Stoichiometric Deficiency The following examples (EX5A-EX5F) show that the electrochemical properties can be further improved when the sample has a slight stoichiometric deficiency of lithium. These samples were made in the same way as sample EX4, but little Li ₂ CO ₃ was added and in some cases the sintering temperature increased slightly.

In many cases, 20 μm of Li ₂ CO ₃ 90% (= REF1) was mixed with jet milled MOOH and Li ₂ CO ₃ 10%. The molar ratio of Li (in Li ₂ CO ₃ ) to MOOH is listed in Table 8 below. Table 8 shows the sintering temperature and describes the measurement result of the BET surface area. In the column of Li: M, the result of the ratio of lithium to the transition metal obtained by chemical analysis of the final sample is described. Chemical analysis results were expected even when sample REF1 had about 1.02 Li: Co and, depending on the temperature, always considered that a small amount of lithium would evaporate during sample manufacture. Very similar to the value. Obviously, samples EX5D, EX5E and EX5F have increasingly lithium stoichiometric deficiencies. SEM analysis is performed and a total of 6 samples are confirmed to show island morphology. An SEM micrograph of sample EX5E is shown in FIG. 10a. X-ray analysis showed a mixture of the two phases in all cases (see below).

  Coin-type batteries were manufactured and tested under the same conditions as described above. The results are summarized in Table 9 below.

Electrochemical data was obtained from two sets of two coin cells. The first set of two cells was tested using a cycle stability schedule. The other set was tested using a rate performance schedule. The cycle stability plan lists the following numbers: Q _rev , Q _irr , fade rate (C / 10) and fade rate (C1) listed in Tables 3 and 9 . The electrochemical data is the average of each set of two batteries. Q _rev and Q _irr are the first cycle reversible discharge capacity (mAh / g) and irreversible discharge capacity (%, Q _irr = [QCh−QDC] / QCh, measured at the C / 10 rate. ). The number of fade rates at C / 10 is obtained by comparing the discharge capacities at the slow (C / 10) 3rd cycle and the 41st cycle, and the fade rate at 1C is fast (1C) 4 It is obtained by comparing the discharge capacities in the first cycle and the 42nd cycle. From the 5th cycle to the 40th cycle, the battery is cycled at 4.5-3.0V with a C / 5 charge rate and C / 2 discharge rate. The fade rate is extrapolated to be 100 cycles.

  The rate performance plan table lists the numbers 1C / 0.1C, 2C / 0.1C, and 3C / 0.1C for the rate performance listed in Tables 3 and 9. The plan is as follows. After the first slow cycle (C / 10), the battery is charged at C / 5 and discharged at an increasing rate (C / 5, C / 2, 1C, 1.5C, 2C, 3C, 5C and 10C). . The voltage range is 4.3 to 3.0V.

In order to reliably measure the discharge capacity and rate performance, the battery electrode loading (g / cm ² ) was different. The cell tested for the stability plan had an electrode loading of about 12 mg / cm ² . Batteries tested against planned rate had a loading of about 5 to 6 mg / cm ^2.

  The data in the table shows that rate performance increases when the Li: M ratio is lowered. The highest rate is obtained for samples having a lithium stoichiometric deficit of about 1.5%. At the same time, 1.5% lithium stoichiometric deficiency sample EX5E also exhibits the highest cycle stability at 4.5V. However, if the lithium stoichiometric deficiency is too great, the properties will deteriorate. That is, sample EX5X, which has a lithium stoichiometric deficit of about 3%, has poor discharge capacity and very poor rate performance.

Crystal Map X-ray diffraction patterns of reference samples REF1, REF2, comparative samples CX2, CX3 and samples EX1-EX3 were obtained. Samples CX2, CX3, EX1 to EX3 are composed of _two phases, a _first phase based on LiCoO ₂ and a second phase based on Li ₁ M′O ₂ . The lattice constants of these phases are obtained by improving the Rietveld of the two phases and can be compared with the lattice constants of the samples REF1 (LiCoO ₂ ) and REF2 (Li ₁ M′O ₂ ) obtained by one phase. it can.

  Table 2 shows the results. FIG. 11 shows the results by a suitable method in which the inventors plot the hexagonal c-axis against the hexagonal a-axis as a crystal map. In this figure, crystal maps of samples REF1, REF2, CX2, CX3, EX1, EX2 and EX3 are described. The entrance shows an enlarged replot of a small area indicated by the rectangle.

Table 2 and FIG. 11 show that the lattice constant of Phase 2 (Li ₁ M′O ₂ ) of Samples EX1, EX2 and EX3 changes significantly and deviates from the value of REF2, which is different from the phase in CX2 and CX3. It is very clearly shown that the lattice constant of 2 is identical to that of REF2. This change is even more pronounced as the sintering temperature increases. When the sintering temperature is increased, causing the movement towards the LiCoO ₂ away from REF2 position map location is expected. This change in position on the map is typical for solid solutions between LiCoO ₂ and Li ₁ M′O ₂ . Apparently, cobalt diffused from Phase 1 (LiCoO ₂ ) into Phase 2 (Li ₁ M′O ₂ ) particles.

Surprisingly, the lattice constant of phase 1 (LiCoO ₂ ) did not change during sintering. All samples CX2, CX3 and EX1, EX2 and EX3 have the same lattice constant as that of REF1.

The Rietveld improvement yields a fraction of phase 2 (Li ₁ M′O ₂ ), which is listed in Table 2. This data shows that the fraction of phase 2 increases during sintering. The fraction of Li ₁ M′O _{2 in} sample CX2 should be 40%. Clearly, the Rietveld improvement describes even higher values for the Li ₁ M′O ₂ phase. This misconception is the rearrangement of small (phase 2, Li ₁ M′O ₂ ) and large (phase 1, LiCoO ₂ ) particles during X-ray sample preparation that may cause enrichment of phase 1 adjacent to the surface. Seems to be caused by. This effect appears to be improved by the preferred orientation of the phase 1 particles. The fraction of Li ₁ M′O ₂ increases with the sintering temperature. It shows that more Co diffuses from Phase 1 (LiCoO ₂ ) to Phase 2 than Ni (and) Mn diffuses from Phase 2 to Phase 1 during sintering.

FIG. 12 shows a crystal map with data points for samples EX4, EX5A-EX5F and CX5 along with samples REF1, REF2. Data points were obtained by improving Rietveld in two phases. This figure clearly shows that the lattice constant of EX4, EX5A-EX5F, phase 2 (Li ₁ M′O ₂ ) exists between the lattice constants of LiCoO ₂ —REF _{1 and} REF ₂ —Li ₁ M′O ₂ . This is consistent with the discussed diffusion of Co into the second phase. At the same time, the lattice constant of phase 1 does not change at all and is the same as that of REF1-LiCoO ₂ .

  FIG. 12 also compares the sample CX5 with samples EX4 and EX5A to EX5F based on positions on the crystal map. This FIG. 12 can include that the lattice constant of phase 2 of CX5 is the same as REF2. This is consistent with the much lower sintering temperature that causes sufficient cation exchange between Phase 1 and Phase 2.

X-ray diffraction patterns Samples REF1 and REF2 have high crystallinity, and therefore these samples exhibit an X-ray diffraction pattern with sharp diffraction peaks. FIG. 13 shows X-ray diffraction patterns (base line: scattering angle (deg)) of CX2, CX4, CX5, and EX1. All the above samples have the same composition. The entrance of FIG. 13 shows an enlarged replot of the region indicated by the rectangle. Sample CX2, which is a mixture of REF1 and REF2 (heat treated), shows an X-ray diffraction pattern that is superposition of REF1 and REF2, as expected. Even if the mixture was heat treated at 900 ° C. (sample CX4) or the mixture of hydroxide and Li ₂ CO ₃ mixed with LiCoO ₂ (CX5) was heat treated at 900 ° C., the X-ray diffraction pattern Remains the same. This teaches us that the first phase LiCoO ₂ and the second phase Li ₁ M′O ₂ did not change.

However, this situation is very different from the sample that is typical for the present invention. FIG. 13 shows that the peak position and peak shape of sample EX1 changed. The peak of phase 1 (based on LiCoO ₂ ) retains a very sharp sample and is identical in position, but the peak of phase 2 (based on Li ₁ M′O ₂ ) is significantly enlarged The position of this peak has clearly moved. The main reason for this expansion is the distribution of stoichiometric amounts of Co and Ni. During sintering, Ni diffuses from the second phase and cobalt diffuses into the second phase. As a result, different particles and / or crystallites have different stoichiometric amounts, and each stoichiometric amount has its own peak position, and as a result, a wider diffraction pattern. A peak is observed. In improving Rietveld, it is difficult to simulate the distribution of lattice constants. Fortunately, however, small crystal dimensions cause some similar peak broadening. That is, the Rietveld improvement of a typical positive electrode of the present invention exhibits a large crystallite dimension relative to the first phase (based on LiCoO ₂ ) and the second phase (based on Li ₁ M′O ₂₎ . And smaller crystallite dimensions. At the same time, the peak position of the diffraction peak of the second phase preferably moved towards the position of the first LiCoO ₂ phase.

  FIG. 14 shows X-ray diffraction patterns of Samples CX6 and EX9E. All the above samples have the same composition. Sample CX6 is different from the above sample. This sample was sintered too strongly. Therefore, the progress of diffusion was extremely remarkable. As a result, the second phase becomes the same as the first phase and can no longer be distinguished by the X-ray diffraction pattern. All the same parts are the small shoulders of the phase 1 peak towards the lower corners. In contrast, sample EX9E shows a small but distinct peak at the lower corners. Some of the peaks in FIG. 14 are indicated by arrows.

In contrast, sample EX9E shows a small but distinct peak at the lower corners. According to our understanding, Co moves into the second phase, the amount of Co increases, and the lattice constant of this Co “moves” towards the lattice constant of LiCoO ₂ (see above). Therefore, the X-ray peaks move closer together, overlap and eventually coincide. Therefore, phase 2 in the oversintered phase does not appear to disappear, but becomes too similar to distinguish from LiCoO ₂ .

It can be concluded that the positive electrode according to the present invention exhibits an X-ray diffraction pattern that can be approximated as LiCoO ₂ with high crystallinity and Li ₁ M′O ₂ with low crystallinity. The crystallinity is still quite good in relation to both phases. Some commercial positive electrode materials are less crystalline than Phase 2. Also, the lattice constant of the second phase is lower than expected (the peak is very close to the LiCoO ₂ peak); the expected value is Li ₁ M′O ₂ prepared from the same MOOH precursor. A typical value for the phase.

  Table 2 summarizes the results of Rietveld improvements.

Voltage profile Coin-type batteries were prepared from all reference samples REF1, REF2, all comparative samples CX2, CX3 and EX1, EX2 and EX3. The voltage profiles of CX2, CX3 and EX1-EX3 during slow discharge are shown in FIG. Samples CX2 and CX3 show a clear plateau at 3.88V. This plateau is typical for LiCoO ₂ . The presence of this plateau indicates that phase 1 is pure LiCoO ₂ . However, in connection with samples EX1, EX2 and EX3, this plateau disappears progressively with increasing sintering temperature. Apparently, phase 1 is no longer LiCoO ₂ . This is consistent with the fact that phase 1 particles contain Ni and Mn, as evidenced by EDS analysis. However, a very Surprisingly, phase 1, having exactly X-ray diffraction pattern of LiCoO _2, and the value expected for LiCoO ₂ that is Ni-Mn-doped with a very different lattice constants.

Rate Performance and Cycle Stability Table 3 is derived from the calculated values for reference examples REF1 and REF2 and coin cell battery tests of samples CX2, CX3, EX1, EX2 and EX3, and hypothetical sample CC1. Results are described. All samples all have the same composition. This table lists the averaged data for two coin cells of each sample.

We observe that sample CX2 (a mixture of heated LiCoO ₂ and Li ₁ M′O ₂ ) has properties very similar to those of the hypothetical sample. Obviously, the mixing of LiCoO ₂ and Li ₁ M′O ₂ no longer yields an advantage. Samples CX3 and CX4 (heated mixtures of LiCoO ₂ and Li ₁ M′O ₂ ) have slightly better rate performance and slightly improved cycle stability, but in general these Is almost the same as the sample CX2 or CC1.

  However, samples EX1, EX2 and EX3 advantageously exhibit improved rate performance. 1C, 2C, 3C compared to 91-93%, 86-88% and 83-86% of hypothetical sample CC1 or mixture CX2 or compared to 94%, 91% and 89% of sample CX3 Then, discharge capacities of 95%, 93% and 91% are obtained.

  We note that the improved rate performance is not related to different forms. All samples CX2, CX3, EX1-EX3 have almost the same BET surface area, and all samples are agglomerated with dense, irregularly shaped large particles (phase 1) A mixture with very small particles (phase 2). Furthermore, the particle size distribution is almost the same. Achieving increased rates without increasing the BET surface area is a critical aspect of the present invention. In principle, it would be possible to reduce the BET surface area to meet safety and compactness requirements and still achieve sufficient rate performance.

  At the same time, the cycle stability of EX1, EX2 and EX3 is dramatically improved. FIG. 16 shows the data obtained for sample EX1. FIG. 16a shows that the fade rate per 100 cycles (discharge capacity vs. cycle number #) is calculated to be 6.4%. A small point represents the electric capacity at the time of charge, and a large point represents the discharge capacity at the time of discharge. FIG. 16b shows the cycle stability of EX1. 16c shows the rate performance of EX1.

  In FIG. 17, the cycle behavior of sample CX6 (left side: A) is compared with EX5E (right side: B).

Claims (25)

A powdered lithium transition metal oxide containing LiCoO ₂ particles supporting Mn and Ni, the particles having islands enriched in Mn and Ni content on the surface of the particles, Said pulverulent lithium transition metal oxide wherein has a higher Mn and Ni concentration than in the body of the particles and the island contains at least 5 mol% Mn. The powdered lithium of claim 1, wherein the island enriched in Mn and Ni has a thickness of at least 100 nm and covers less than 70% of the surface of the LiCoO ₂ particle phase supporting Mn and Ni. Transition metal oxide. The powdered lithium transition metal oxide according to claim 1 or 2, wherein the Mn concentration in the island is at least 4 mol% higher than the Mn concentration in the main body of the LiCoO ₂ particle phase supporting Mn and Ni. Ni concentration in islands content of the Mn and Ni were enriched at least 2 mol% higher than Ni concentration in the body of LiCoO ₂ particle phase supporting the Mn and Ni, one of the Claims 1 to 3 2. The powdery lithium transition metal oxide according to claim 1. The powdery lithium transition metal oxide according to any one of claims 1 to 4, wherein the LiCoO ₂ particles supporting Mn and Ni contain at least 3 mol% of both Ni and Mn. The crystal lattice constants a and c of LiCoO ₂ particles supporting Mn and Ni are 2.815 +/- 0.002 and 14.05 +/- 0.01, respectively. 2. Powdered lithium transition metal oxide according to item 1. The powdered lithium transition metal oxide according to any one of claims 1 to 6, wherein the LiCoO ₂ particles supporting Mn and Ni are monolithic and have no internal porosity. The powdery lithium transition metal oxide according to any one of claims 1 to 7, wherein the particle size distribution of LiCoO ₂ particles supporting Mn and Ni has a d ₅₀ of _more than 10 µm. The powdery lithium transition metal oxide according to any one of claims 1 to 8, comprising LiCoO ₂ particles supporting Mn and Ni in an amount of 30% by mass to 95% by mass. A first phase composed of LiCoO ₂ particles supporting Mn and Ni and further _a second island-free phase of the general formula Li _{1 + a} M ′ _1-a O _{2 ± b} , -0.03 <a <0.05 and b <0.02, M 'is _{Ni m Mn n Co 1-mn} , is m> n and 0.1 <m + n <0.9, claims The powdery lithium transition metal oxide according to any one of 1 to 9. Li _x M _y O ₂ having an overall composition of _{± [delta],} in this case 0.97 <x <1.03,0.97 <y < 1.03, x + y is at 2; δ <0.05; and M is _{Co 1-fg Ni f Mn g} , 0.05 < a f + g <0.5 and f ≧ g, powdered lithium transition metal oxide according to any one of claims 1 to 10 object. The powdered lithium according to claim 11 , wherein the oxide is composed of two phases, the first phase is LiCoO ₂ particles supporting Mn and Ni, and the second phase is an island-free phase. Transition metal oxide. Islands-free phase of the crystal lattice constant a 'and c', the same composition Li _x M _y O 2 has a _{± [delta]} and pure LiCoO ₂ references lithium particles are composed of island-free phase and the corresponding transition The following relationship with the lattice constants a ″ and c ″ of the corresponding island-free phase of the metal (M _ref ) oxide:
13. The powdered lithium transition metal oxide of claim 12, having 0.980 <a ′ / a ″ <0.998 and 0.9860 <c ′ / c ″ <0.9985. Island-free phase has a second particles having a particle size distribution with a d ₅₀ of 2 to 10 [mu] m, the second particles, a primary having a particle size distribution with a d ₅₀ of 0.5~2μm The powdery lithium transition metal oxide according to any one of claims 10 to 13, which is composed of a sintered aggregate of crystallites. Furthermore both the island and the island-free phase the content of Mn and Ni-enriched, containing Ti, Ti content in this case is less than M10 mol% in oxide _{Li x M y O 2 ± δ} , The powdery lithium transition metal oxide according to any one of claims 11 to 14. Further, the oxide Li _x M _y O ₂ in _{± [delta],} 5 mol% and less than Be the M of one or more dopants selected from the herd that is configured of Al and Mg, B, Ca, Zr, The powdered lithium transition according to any one of claims 11 to 15, containing less than 1 mol% of said M of one or more dopants selected from the group consisting of S, F and P. Metal oxide.   The electrochemical cell which has a positive electrode containing the powdery lithium transition metal oxide of any one of Claim 1-16 as an active material. Next step:
Preparing a mixture of LiCoO ₂ powder or a cobalt-containing precursor compound having a cobalt content of at least 90 mol% and Li-Ni-Mn-Co oxide or Ni-Mn-Co precursor powder and Li precursor compound;
The mixture is sintered at a temperature T of at least 910 ° C. for a time t of 1 to 48 hours,
17. The method according to claim 1, further comprising obtaining LiCoO ₂ particles supporting Mn and Ni having islands enriched with Mn and Ni on the surface. Of producing a powdered lithium transition metal oxide. The Ni-Mn-Co precursor powder is a transition metal hydroxide, transition metal oxyhydroxide, transition metal carbonate, transition metal oxycarbonate, or lithium transition metal compound, in which case the transition metal composition M '' the method of, M '' = a _{Ni o Mn p Co 1-op} , which in this case o + p> 0.5, and o> p, claim 18.   The method according to claim 18 or 19, wherein the Ni-Mn-Co precursor powder contains a transition metal content of the powdered lithium transition metal oxide of 5 to 70 mol%. LiCoO ₂ powder has a tap density of at least 2 g / cm ^3, and is composed of a monolithic particles having a d ₅₀ of at least 10 [mu] m, method according to any one of claims 18 to 20.   21. A method according to any one of claims 18 to 20, wherein the cobalt-containing precursor compound is one of cobalt hydroxide, cobalt oxyhydroxide or cobalt carbonate. The LiCoO ₂ or cobalt-containing precursor contains at least 80% of the transition metal of the powdered lithium transition metal oxide, and the Ni—Mn—Co-containing precursor powder has a particle size distribution with a d _{50 of 1} to 3 μm. 23. A method according to any one of claims 18 to 22 comprising particles. The LiCoO ₂ or cobalt containing precursor contains at least 80% of the transition metal of the powdered lithium transition metal oxide, and the Ni—Mn—Co containing precursor powder has a particle size distribution with a d ₅₀ of 4-10 μm. 23. A method according to any one of claims 18 to 22, wherein the method consists of agglomerated particles. Further, Ni-Mn-Co-containing precursor has the form of TiO ₂ particles having a d ₅₀ of less than 100nm and Ti, method according to any one of claims 18 to 24.

Images