Strategies and Perspectives to Catch the Missing Pieces in Energy‐Efficient Hydrogen Evolution Reaction in Alkaline Media

Abstract Transition metal hydroxides (M‐OH) and their heterostructures (X|M‐OH, where X can be a metal, metal oxide, metal chalcogenide, metal phosphide, etc.) have recently emerged as highly active electrocatalysts for hydrogen evolution reaction (HER) of alkaline water electrolysis. Lattice hydroxide anions in metal hydroxides are primarily responsible for observing such an enhanced HER activity in alkali that facilitate water dissociation and assist the first step, the hydrogen adsorption. Unfortunately, their poor electronic conductivity had been an issue of concern that significantly lowered its activity. Interesting advancements were made when heterostructured hydroxide materials with a metallic and or a semiconducting phase were found to overcome this pitfall. However, in the midst of recently evolving metal chalcogenide and phosphide based HER catalysts, significant developments made in the field of metal hydroxides and their heterostructures catalysed alkaline HER and their superiority have unfortunately been given negligible attention. This review, unlike others, begins with the question of why alkaline HER is difficult and will take the reader through evaluation perspectives, trends in metals hydroxides and their heterostructures catalysed HER, an understanding of how alkaline HER works on different interfaces, what must be the research directions of this field in near future, and eventually summarizes why metal hydroxides and their heterostructures are inevitable for energy‐efficient alkaline HER.


Introduction
Eco-friendlier and efficient hydrogen production has been one of the most attention-grabbing topics of energy research ever since the idea of hydrogen economy was coined with avision to achieve the "affordable and clean energy" goal of 17 sustainable development goals proposed by the united nations. [1,2] Current cost-efficient method, the steam reforming of methane provides affordable hydrogen at the cost of low purity. [3][4][5][6] Besides,i ta lso requires methane/other low molecular weight hydrocarbons that are quickly depleting or more precisely being overconsumed with an exponentially increasing rate.T his suggests that steam reforming thereby cannot be av iable option for all our future hydrogen demand. [7,8] As imilar method that primarily consists of steam reforming for hydrogen production is biomass gasification and catalytic reforming under steam. [9,10] Though this method seems to be apractical option as biomass production is anticipated to increase linearly with the increasing population, the amount of hydrogen that can be produced by this method will never be sufficient to meet the demand in the future.Biomass gasification is carbon-neutral, however,both steam reforming and biomass gasification release greenhouse gases (GHGs) and hence,t hey are not environmentally friendlier. [11,12] Because of the aforementioned disadvantages albeit having cost-efficiencies,s team reforming of methane and biomass gasification are to be replaced with an advantageous method that produces hydrogen with the highest purity Transition metal hydroxides (M-OH) and their heterostructures (X j M-OH, where Xcan be ametal, metal oxide,metal chalcogenide, metal phosphide,e tc.) have recently emerged as highly active electrocatalysts for hydrogen evolution reaction (HER) of alkaline water electrolysis.L attice hydroxide anions in metal hydroxides are primarily responsible for observing suchanenhanced HER activity in alkali that facilitate water dissociation and assist the first step,t he hydrogen adsorption. Unfortunately,their poor electronic conductivity had been an issue of concern that significantly lowered its activity. Interesting advancements were made when heterostructured hydroxide materials with ametallic and or asemiconducting phase were found to overcome this pitfall. However, in the midst of recently evolving metal chalcogenide and phosphide based HER catalysts,s ignificant developments made in the field of metal hydroxides and their heterostructures catalysed alkaline HER and their superiority have unfortunately been given negligible attention. This review,u nlike others,b egins with the question of why alkaline HER is difficult and will take the reader through evaluation perspectives,t rends in metals hydroxides and their heterostructures catalysed HER, an understanding of how alkaline HER works on different interfaces,what must be the researchd irections of this field in near future,and eventually summarizes why metal hydroxides and their heterostructures are inevitable for energy-efficient alkaline HER. and emits no GHGs.Amethod that fits within these criteria is electrocatalytic water splitting. [13][14][15][16][17][18][19] Electrocatalytic water splitting is regarded as the fastest, safest, and the greenest (provided that input energy is from renewables) method of all capable of producing highly pure hydrogen (99.999 %) while also being regarded as an indirect mean of large-scale energy storage that stores electrical energy as chemical fuels. [20,21] In spite of having such advantages,w ater electrosplitting lacks in energy-efficiency and also suffers from avery low abundance of the best active catalysts which have been the main reasons for hindering its successful commercialization. [22][23][24] Affordability of hydrogen produced during catalytic water electrosplitting is determined primarily by the activity and the availability of electrode materials used. [25,26] Materials that perform the half-cell reactions (oxygen evolution reaction (OER) at the anode and hydrogen evolution reaction (HER) at the cathode) with ab etter energy efficiency (i.e., with lower overpotentials) involve Pt (for HER) and the oxides of Ir and Ru (for OER). These metals are very low in abundance that forbids the successful scale-up of water electrolysers for large-scale hydrogen production. [25,26] To avoid this issue with these precious metals based electrocatalysts,several strategies were proposed and noteworthy ones are nanostructuring and noble metal dilution by alloying and soft-templating in colloidal solutions. [27][28][29][30][31] Energy-efficiencyo fh ydrogen production is also improved by performing the electrolysis with aqueous solutions of extremely low or high pH (such as 0( e.g. 0.5 M H 2 SO 4 or 1.0 MH ClO 4 )a nd 14 (e.g. 1.0 MK OH)) and the recent developments in the area of proton exchange membrane (PEM) for acidic water electrolysis and alkaline water electrolysis equipped with anion exchange membranes (AEM) are significant in the past two decades. [20,32,33] Besides, aremarkable amount of research works is also being done in the area of neutral and near-neutral electrocatalytic water splitting with av ision of improving the efficiencyo fs olar to fuel conversion (SFC) devices,the activity of which is largely dependent on the electrocatalytic activity of materials used in neutral and near-neutral waters. [13,[34][35][36][37][38] Acidic water electrolysis is blessed with an extensive library of recently evolved non-precious HER catalysts that perform better than Pt in acid. [7,8] However,itstill lacks in the development of highly active and reasonably stable OER catalysts made of cheaper materials.K nown acidic OER catalysts of cheaper materials often contain the oxides of Co and Mn with other post-transition elements that fetch the ability to self-heal under ah ighly corrosive acidic environment under oxidizing potentials. [39][40][41] Alkaline water electrolysis,i nc ontrast, is blessed with av ast variety of non-precious electrocatalysts for OER that include hydroxides/oxyhydroxides, oxides,c halcogenides, pnictides,c arbides,b orides and intermetallics. [17,28,[42][43][44][45] Among them, those non-oxide/hydroxide catalysts are recently comprehended to act as precatalysts rather than actual OER catalysts because of the lability of non-oxide/ hydroxide anions that are easily displaced by hydroxide anion under highly alkaline solution. [46][47][48][49] Such ad isplacement of non-oxide/hydroxide anions occurs even under reductive potentials in alkaline solutions. [50,51] Because of this phenomenon, it is now understood that most of the non-oxide/ hydroxide electrocatalysts that have so far been reported for alkaline HER must have catalysed it by forming am etal hydroxide phase on the surface that acted as the real-time catalyst. Hence,i ti se xplicit that for efficient alkaline HER, am etal hydroxide surface of an appropriate electronic configuration and moderate hydridic bond strength is inevitable. [50,52,53] Thepresence of lattice hydroxide anions is likely to enhance the water dissociation by facilitating the initial water adsorption with the partial negative charge that it possesses by hydrogen bonding. [50,52,53] Similarly,arelatively better HER performance witnessed with non-oxide/hydroxide catalysts are likely due to the improved conductivity imparted by the conducting metallic and/or semiconducting phosphide/chalcogenide/carbide phases that also provide highly active defect sites that are formed by lattice mismatches at the junction of hydroxide phase and the heterophase which can be of several kinds in nature (Scheme 1). [53] Perceiving the significance of metal hydroxides and their heterostructures (doped hydroxides,c ore-shell hydroxides, and hydroxides with slabs of hetero-phases) catalysed alkaline HER in improving the energy and cost-efficiencyo f alkaline water electrolysers,w eb elieve that these catalysts are the missing pieces of the puzzle (i.e., energy-efficient alkaline HER) that we have been dealing with. Further understanding and development in this area of electrocatalysis research, therefore,w ill ensure ac omparable energy efficiency to that of PEM water electrolysers but fascinatingly with non-precious catalysts which are and will be ag reat breakthrough in the field of water splitting electrocatalysis.

Alkaline HER:W hy Is It So Hard?
Evolution of hydrogen gas from an aqueous electrolyte is largely pH dependent and is facilitated in solutions with avery high proton activity (i.e., acid solutions with pH 0). In such ah ighly acidic solution, regardless of the nature of the surface (metallic or semiconducting) that catalyse HER under applied potential, HER can proceed via two different paths namely Volmer-Heyrovský mechanism and Volmer-Tafel mechanism. In both paths,V olmer step is inevitable. Each step involved in acidic HER is given below [Eqs.(1)- (3)]. [54] Vomer step : H þ ðaqÞ þ e À ! H ads ð1Þ As can be witnessed, proton adsorption and discharge that forms adsorbed Hatoms (H ads )would be facile in asolution of extremely low pH. Thefollowing step could either be Tafel or Volmer depending on the work function, density of states,and the local crystallographic geometry of the catalytic sites. [55,56] In general, close-packed systems with higher probabilities of forming more H ads in closer proximities tend to evolve hydrogen following Tafel path and this is the most efficient path for HER that require minimum work (i.e., less applied potential). [18] On the other hand, catalysts with widely dispersed catalytic sites in which the distance between two adjacent catalytic sites that could form H ads is larger than the sum of Va nder Waals radii of two SÀH ads bonds (S represents the catalyst site) tend to follow the Heyrovský step instead. [8] This is an energy-inefficient path when compared to the Volmer-Tafel mechanism. Hence,itisalways desired to have HER catalysts that would follow Volmer-Tafel mechanism. A simple steady state polarization curve analysis that helps us to calculate the Tafel slope will tell us the mechanism. [57] Catalysts that have Tafel slopes closer to 30 mV dec À1 are said to follow Volmer-Tafel mechanisms and others with Tafel slopes higher than 45 mV dec À1 are said to preferably follow the Volmer-Heyrovský mechanism. Thes electivity between these two mechanisms shifts towards Volmer-Heyrovský mechanism with the increasing Tafel slope until 120 mV dec À1 .I na cidic conditions,i ti sv ery rare to see aH ER electrocatalyst that could have an abnormal Tafel slope (> 120 mV dec À1 )a st he solution is rich in protons and the Volmer step is not hindered/dependent by any other reaction that must occur along.N onetheless,i ti sa lso to be emphasized here that the Tafel analysis of HER in acidic conditions can be deceiving as proton discharge and the delivery of H 2 molecule are very fast in proton-rich solutions. Therefore,t he measured Tafel slopes often reflect the mass transport limitations rather than reflecting the mechanism by which HER is being catalysed.
In alkaline conditions where the probability of finding af ree proton in solution is nearly zero and the Volmer step (proton adsorption and discharge) has to rely always on aw ater dissociation reaction which is thought to be accom- panied by the Volmer reaction and occurs separately prior to the Volmer reaction in the vicinity of electrode surface according to the beliefs of another part of researchers. [50,53] Irrespective of when and how water dissociation occurs,t he Volmer step of HER in alkaline solution is not identical to that of the one given in Equation (1) and so is the Heyrovský step as they both depend on water dissociation to form H ads / H 2(g) .Here,the three steps of HER in alkali are listed below [Eqs.(4)- (6)].
Volmer step À H 2 ODissociation : From the Volmer and Heyrovský steps of alkaline HER, it can be noted that the interface is given an extra burden to carry which is the water dissociation step providing the required free proton for the reaction to proceed. This implies that HER in alkali is indeed harder than acidic HER. In addition, it could also be noted that ah ydroxide anion is produced locally at the surface of the electrode with the expense of aw ater molecule,w hich further makes things complicated as it will lower the concentration of water molecules and pushes the equilibrium of water dissociation back. These insights from the mechanism suggest that alkaline HER is an energy intensive process than acidic HER and even the state-of-the-art Pt struggles to perform efficient alkaline HER. [58] Figure 1, ab ar diagram showing the values of exchange current density (j 0 )o fH ER/hydrogen oxidation reaction (HOR), Tafel slopes,and enthalpy changes (DH)o f activation for aPt (111) surface reported by Markovic and coworkers. [59,60] As one can clearly see from Figure 1that Pt(111) surface required higher enthalpy change for activation in alkali than in acid. Similarly,amassive increase in Tafel slope suggests that there was achange of mechanism of HER. ATafel slope of 74 mV dec À1 implies that Pt(111) surface was following Volmer-Heyrovský mechanism in acid whereas an abnormal Tafel slope of 150 mV dec À1 imply that the Volmer step of HER is dependent on another reaction (water dissociation in this case) for it to proceed via proton adsorption and discharge.Inalkaline conditions,itisvery common to witness Scheme 1. Graphicald epiction of metal hydroxides and different types of heterostructured metal hydroxides used in alkaline HER.

Angewandte Chemie
Reviews such high and abnormal Tafel slope values (> 120 mV dec À1 ) that primarily indicate that the reaction is controlled by the water dissociation step and the rate determining step (RDS) is the first electron transfer reaction (i.e., the Volmer step). Hence,any HER electrocatalyst that could have aT afel slope value lesser than 120 mV dec À1 is considered to be better in alkali and the catalysts that could follow the Volmer-Tafel mechanism in alkali are rare which often contain Ru or Pt with metal hydroxide or heterostructured metal hydroxide phases. [61][62][63] Ru doped alkaline HER catalysts are thought to experience an enhancement in activity by the high affinity of Ru to dissociative complexation with water molecules that provides the required localized proton reservoir. [63] On the other side,m etal hydroxide phases with lattice hydroxide anion that are relatively deficient in electron population (as they are in the coordination sphere of metal ions) when compared to free hydroxide anion in the electrolyte are less likely to repel water molecules away and in fact, they are believed to attract water molecules closer to electrode surface via hydrogen bonding.B ecause of these facts,m etal hydroxides,heterostructured metal hydroxides and metal hydroxides doped with Ru are performing better in alkali, particularly, the latter one outperformed Pt in alkaline HER. [53] Hence,it is now explicit that for efficient alkaline HER, the metal hydroxide phase is inevitable.

Evaluation Perspectives:H ow Are HER Electrocatalysts Screened and Benchmarked?
As with all electrocatalysts,aHER electrocatalyst is also evaluated for high activity,s electivity,a nd stability which in combination with high abundance make it aperfect choice for energy and cost-efficient hydrogen production. [54,64] Activity is determined by both thermodynamics and kinetics of HER on ag iven electrocatalystss urface.I ng eneral, an electro-catalyst is expected to catalyse the reaction of interest at the thermodynamically determined reversible/equilibrium potential (E 0 )under ideal conditions.However,inreal cases,there are other phenomena (mainly the kinetic complexities such as the water dissociation step in alkaline HER) that add further workload and makes the electrode to demand ac ertain quantity of additional energy (i.e.additional applied potential familiarly known as onset overpotential) to begin the reaction of interest. Beyond this onset overpotential, for every increase required in terms of current density will also increase the overpotential. Hence,i ti sd esired to have ac atalyst that could have lesser onset potential. An important characteristic of this onset overpotential is that it is intrinsic to the catalysts surface properties and does not vary with the varying loading/ specific surface area (SSA) of an electrocatalyst. This characteristic of onset overpotential places it among other important activity markers of aH ER electrocatalyst. Nonetheless,t here are several examples of electrocatalysts which were shown to have relatively higher onset overpotentials yet deliver better activity at higher overpotentials (current densities) due to various reasons among which increased SSA (thus electrochemical surface area (ECSA)) as aresult of nanostructuring and increased loading are being the primary contributors to the overall enhancement witnessed. [65][66][67] An example is NiTe 2 nanowires electrocatalysed HER in acid and alkali (Figure 2a,b) which apparently required higher onset overpotential than Pt/C but outperformed the latter at higher overpotentials in current density. [66] This implies that relying only on onset overpotential for determining the activity of an electrocatalyst could either overrate or underrate its actual performance at higher applied overpotentials,w hich is the case in real water electrolysers.H ence,r esearchers have adapted the practice of benchmarking these electrocatalysts at 10 mA cm À2 in analogy to the solar to fuel conversion (SFC) devices evaluation. [64,[68][69][70] Forh igh-performance catalysts, overpotentials at 100, and 500 mA cm À2 are also often reported along with onset overpotential and the overpotential at 10 mA cm À2 .H owever,t he recent comprehension is that any catalyst that is capable of delivering 1mAcm À2 at lower overpotentials are considered better and they can be scaled up to improve their activity extrinsically depending on the requisites.T hough onset overpotential and overpotential at any fixed current density (in cathodic direction) are primarily controlled and dictated by the hydrogen overpotential deposition (H opd ), hydrogen underpotential deposition (H upd )was earlier thought to play crucial roles in determining the HER/HOR activity of electrocatalysts in water electrolysers and fuel cells.Hydrogen binding energy (HBE) of H upd is highly pH dependent and it shifts toward more negative values (i.e., the corresponding potential will shift anodically) with increasing pH as reported by Yanand co-workers for Pt surfaces (Figure 2c). [71] By this study,they revealed that such ac hange in HBE towards more negative values (i.e., the anodic shift of H upd potential) is inversely related to the HER activity of Pt surfaces they studied. Laying the foundation on this finding, Zheng and co-workers [72] later came up with am ore convincing way of determining HER activity of Pt group metals (Ir, Pd, Pt, and Rh) supported on conducting carbon by plotting the exchange current density (j 0 ,t he  [59,60] current density measured for HER at reversible overpotential) against H upd desorption peak (Figure 2d). From their study,itwas found that catalysts with higher j 0 and lower H upd desorption potential (indicating weaker bonding of hydrogen on catalystssurface) are highly active for HER. As indicated earlier, the hydrogen desorption/adsorption potential was found to shift anodically with increasing pH which justifies poor HER activity of electrocatalysts in highly alkaline  (100) surfaces. [71] (d) Alinear relationship arrived between the natural log of exchange current density and H upd desorption peak for Pt group metals. [72] (e) The plot of rate of HER of various metals decorated on Pt(553) single crystal electrode in alkali against the DFT-calculated hydroxide binding strength at 0.0 Vvs. RHE. (f)The 3D volcano relationship among the binding strengthso fhydrogen and hydroxide ion with the energy of water dissociation for the single crystals Pt (111) and Pt(553), Ru decorated single-crystalPt (111) and Pt(553), and the alloy PtRu (111) showing that aweaker DG H* and astronger DG OH* (at and around À0.3 eV) are essential for better water dissociation in alkaline HER. Both (e) and (f)are reproduced from ref. [74] (Copyright 2020, NPG).
solutions.D espite being an accurate activity marker,t his relationship between j 0 and H upd desorption potential cannot be made for all electrocatalysts especially for semiconducting surfaces that hardly show such H upd characteristics.This made this activity marker inapplicable for most of the recently evolved HER electrocatalysts.Besides,the HBE proposed by these studies is in fact inaccurate in alkaline solutions.T his is because HER is pH dependent like H upd but not Nernstian in nature like H upd .A sar esult, the guesses that could be made (for designing new catalysts) only from HBE may not be as impeccable as it was once expected to be in predicting the activity trends for alkaline HER. Forexample,the HBE was calculated to become highly negative for Pt with the increasing pH. However,t he loss of HER activity by 2to3folds in alkaline solutions was not justified. [73] McCrum and Koper [74] have very recently shown the importance of hydroxide binding strength (DG OH* )a tt he potential 0.0 Vv s. RHE along with HBE for ac atalyst to have af acile water dissociation step in alkaline HER. In this study,m etals such as Mo,Re, Ru, Rh, and Ag were decorated on aPt(553) single crystal electrode and their HER activity in alkali was screened. When the rate of HER was plotted against the DG OH* ,itled to aclassical Volcano relationship ( Figure 2e)in which Ru decorated Pt(553) occupied the atop position with the highest rate of HER and astronger DG OH* .This study has proven that strength of both hydrogen binding and hydroxide binding are to be optimized for better water dissociation and an efficient HER activity in alkali. Specifically,acatalyst that binds hydrogen weakly and hydroxide strongly (at and around 0.3 eV) can act better in water dissociation step which is the bottleneck of efficient HER in alkali ( Figure 2f). Besides, overpotential at geometrical surface area normalized current densities,overpotentials at adefined mass normalized current density (mass activity), ECSA normalized current density (specific activity) are also used to precisely screen the catalyst under examination. Adetailed discussion can be found in our earlier perspectives and reviews. [8,54] All those overpotentials mentioned above are thermodynamic parameters.O nt he other hand, activity is also controlled by the kinetic parameters.T wo other main kinetic parameters that researchers use in HER electrocatalysis are Tafel slope and turnover frequency( TOF) in which the former one qualitatively suggests how faster the HER is occurring on the interface with certain values that are often closer to 30, 60, 90, and 120 mV dec À1 . [54] As far as the Tafel slope is concerned, the lower the value,the better the kinetics. Tafel slopes are inversely related to the charge transfer coefficient of the reaction which is ad irect measure of how efficiently electrons are transferred across the interface. Besides,T afel slopes also serve as am eans to predict the mechanism that the interface follows.A ny value closer to 30 mV dec À1 indicates that the interface catalyses HER by following Volmer-Tafel mechanism whereas values higher than 60 mV dec À1 indicate that the interface is following the Volmer-Heyrovský mechanism. Thes econd kinetic activity marker,T OF is determined using the Equation (7) where j represents current density at ad efined potential, N A stands for Avogadro number, n indicates the number electrons transferred per molecule of hydrogen evolved (two in this case), F denotes the Faraday constant, and G indicates the number of active sites.H ence,t he calculated TOF is expressed as the amount of hydrogen produced per second.
Thus TOFsimply tells us how quick an interface can be in catalysing HER under applied potential and serves as an important activity marker.
Selectivity and stability are two other important perspectives of HER electrocatalyst evaluation that determines the fate of the electrocatalyst under study in promoting it to largescale water electrolysers.S electivity is defined as the percentage efficiency of an electrocatalyst in using the applied energy (potential/current) selectively for the desired reaction (HER in this case) which is commonly termed as Faradaic efficiency/coulombic efficiency. [13,14] Themost efficient way in determining coulombic efficiency of HER electrocatalyst is using gas chromatography (GC). Specifically,t he amount of hydrogen produced under an applied potential is quantified using GC at regular intervals until five to six measurements are made.T hen, these values are compared to the amount of hydrogen that is calculated theoretically using Faradaysl aw of electrolysis.Most of the reported HER electrocatalysts are shown to possess 100 %c oulombic efficiency as there occur no competing reactions usually.Finally,the stability of aHER electrocatalyst is determined classically by two ways in which the first one is the static method. In this method, either ac onstant potential (chronoamperometry) or ac onstant current (chronopotentiometry) is applied and the change in current or potential is monitored with respect to time. [70] This study is carried out at least for 12 hcontinuously and some do perform it for days.D uring the course of this study, degradation in activity is reported in percentage loss.T he second method of stability evaluation is dynamic where ab road potential window covering various electrochemical features of the catalytic electrode including the HER region is chosen and acyclic potential ramp is applied with avery high scan rate such as 100 or 200 mV s À1 and the number cycles may vary from afew hundred to thousands. [14,29,54,75] Any change in onset overpotentials and in overpotentials at defined current densities is measured and reported. An ideal HER electrocatalyst is anticipated to show negligible changes upon such cycling. Therefore,agood HER electrocatalyst should have al ower onset overpotential, al ower overpotential at benchmarking current densities,alower H upd desorption potential (if the interface is ametal), and alower Tafel slope.Similarly, exchange current density (j 0 ), TOF, and coulombic efficiency should be higher. Meanwhile,the effects of substrate electrodesd imension and conductivity and the ways in which the catalyst is supported (i.e., drop-casting,spin-coating,sputtering, thermal evaporation, electrodeposition, hydro/solvothermal solution growth) should not be left aside.D epending on the conductivity and geometry of the substrate electrode,the observed activity of the same electrocatalyst can vary significantly.S uch variations are highly pronounced particularly when the use of different substrate electrodes causes uncertainties in the catalystsl oading. Also,f oam and fiber type 3D substrate electrodes often get access to the electro-lyte beyond what is actually exposed via the capillary action. An elaborate discussion on the mesoscopic effects of substrate electrode geometry can be found in our earlier perspective, [54] the review of Feng and co-workers, [76] and the viewpoint of Zheng and co-workers. [77]

Alkaline HER and Metal Hydroxide Electrocatalysts
As stated earlier, metal hydroxides have recently been found to present at least on the surface of every non-noble metals-based HER electrocatalysts in alkali as aconsequence of displacing action of highly electronegative,n ucleophilic, and arelatively strong field ligand hydroxide. [46,47,78,79] Thus it has now been doubted that each HER catalyst (except Pt, Au, and Hg) ever reported for an alkaline solution must have catalysed it by forming as econdary hydroxide phase at least on the surface.Perceiving this fact, we discuss here the critical advancements made in this field in arriving at such an understanding.U pcoming discussion will give ad etailed account on the application of various metal hydroxides and metal hydroxide heterostructures for alkaline HER. Subsequently,t heir activities in terms of onset overpotential, exchange current density,T afel slope,o verpotential at 10 mA cm À2 ,and TOFare benchmarked while also indicating loading of catalysts which play crucial roles in altering overpotential at 10 mA cm À2 and TOFatt he end.

Metal Hydroxides
Alkaline HER electrocatalysts that contain mainly the hydroxide phase are frequently reported with commendable catalytic activity (see the benchmarking tables). These pristine metal hydroxides can have different crystallographic structures such as layered double hydroxides (LDHs), bhydroxides with a hcp arrangement, and multi-metallic LDHs which is also called layered triple hydroxides (LTHs) when there are three different metals.H owever,L THs vary in no crystallographic aspects when compared to LDHs.Nickel and cobalt hydroxides have also been shown to have appreciable HER activity in the alkali of which Ni(OH) 2 always outperforms Co(OH) 2 with as ignificantly higher margin in overpotentials.Avery first and ad etailed account on the HER (also on the OER activity) activity of Ni(OH) 2 was reported by Corrigan and Bendert in the late 80s. [80] This study revealed the effect of coprecipitating several rare earth, transition, and post-transition metal ions (Cd, Ce,C o, Cu, Fe,L a, Pb,M g, Mn, Ag, Y, and Zn) with Ni(OH) 2 on HER activity (Figure 3a-c). Their key finding was that only Ag and Pb showed strong catalytic enhancement and Ce,C u, and Zn showed ac onsiderable catalytic enhancement in HER activity while the inclusion of other metal ions (Cd, Co,Fe, Y, La, and Mn) did not impart any notable changes in their HER overpotentials at 16 mA cm À2 (Figure 4).
Interestingly,c oprecipitated Cr had poisoned the HER activity of Ni(OH) 2 which is justified by its strong affinity towards forming the structurally rigid trivalent hydroxide. Chemistry of Pb and Ag are almost similar, they both have asimilar density of states,both get precipitated with hydrogen sulphide,b oth form black oxides upon exposure to atmospheric oxygen in the presence of moisture.I nc ontrast, none of the other metal ions studied along with these two have such similarities in their chemistry.Since then, advancements made with pristine hydroxide materials based HER electrocatalysts for alkaline electrolytes remained underdeveloped until the transfiguring nanostructuring of multimetallic hydroxide came into play.A tt he beginning of this decade,t here was ar enewed interest in re-examining first-row transition metal hydroxide for alkaline HER electrocatalysis.H all and coworkers [81] in their study revealed that polished Ni electrodes kept at potentials corresponding to the formation of Ni(OH) 2 for alonger time period gradually increased the HER activity which proved once again that ah ydroxide phase of Ni is better active in alkaline HER than the fresh metallic surface of the same.
Later, hydroxides that had two or more metals were found to perform better than these monometallic hydroxides as they offer av ariety of active sites with different but favourable hydroxide/water adsorption energies.T his helped these catalysts to lower the threshold of Volmer step in alkaline HER. [52] When binary metal hydroxides are considered, LDHs are the ones reported in the highest number for HER in alkali. So far, the reported combinations are Ni-Co, Ni-Fe, Co-Fe, Ni-Cr, and La-Ni. In that series,B aranton and Coutanceu [82] reported av ery systematic study taking Ni and Co binary hydroxide system in which they varied the ratio of Ni and Co as 9:1, 7:3. 5:5, 3:7, and 1:9, respectively,and also compared the activity of only Ni and Co hydroxide.T hey made two significant observations viz.,t he Ni 2+ /Ni 3+ redox couple was shifted cathodically with the increasing Co content and the same had also lowered the HER overpotential. Unexpectedly,o nly Co(OH) 2 delivered better activity than both pure Ni(OH) 2 and all other combinations of Ni and Co in the binary hydroxide they prepared. Despite the fact that ab inary metal hydroxide with an optimal composition must show better HER performance than their monometallic counterparts,t his observation stood out of the order.A careful observation on mass loading that they determined from TGA analysis showed clearly that Co was higher in loading (by weight percentage) than all other catalysts.T his could be the reason for observing such as tandout result. However,t heir study showed that Ni-Co hydroxides of ratio 1:1a nd 1:0.4 provided ab etter compromise between activity and stability.F ollowing this study,B ai and co-workers [83] fabricated aC o-Ni hydroxide composite on a3 DN if oam that the benchmarking 10 mA cm À2 at just 77 mV from the reversible potential of hydrogen evolution. However,such an improved performance was obviously due to the 3D topology and huge mass loading of the catalyst which in turn was reflected by the huge double layer capacitance that they observed when compared to other catalytic interfaces taken along. As imilar Ni-Co LDH fabricated on Ni foam with arelatively lesser loading by Liu and co-workers [84] delivered poor activity by requiring 166 mV as overpotential for the same 10 mA cm À2 .These two contradicting reports testify that besides intrinsic activity changes,m ass loading, specific/ ECSA, and mesoscopic effects play vital roles in determining the HER performances of these binary metal hydroxides. Zhao and co-workers [85] were the first ones to establish the mechanism by which Ni-Co hydroxides perform HER in alkali. They termed it as the "pushing mechanism". According to their study,t he introduction of 10 %C oi nN i(OH) 2 improves the HER performance by increasing the electron density at Ni centres via pushing the electrons that introduced Co had away from it when polarized cathodically.T his pushing effect eased the Volmer step of HER and led to improved performance.Onthe other hand, when they added 10 %o fF ewhich longs for electron under applied potentials decreased the activity.
Interestingly when both Co and Fe were added (20 % cumulatively) to Ni(OH) 2 ,t he HER performance was even worsened. These observations showed that ab inary phase with as uitable metal with an appropriate electron donating ability is essential to observe asignificant HER enhancement. Their observations on HER overpotential, Tafel slope,a nd double layer-capacitance of Ni(OH) 2 with different proportions of Co and Fe have indicated that Ni 0.9 Co 0.1 hydroxide was the optimal system (Figure 5a-f). Though the abovediscussed Ni-Co binary hydroxide systems showed that the addition of Fe into Ni(OH) 2 lattice would lower the HER performance,o ther researchers have shown contradicting results that are discussed below.O ne of such interesting results was shown by Rajeshkhanna and co-workers [86] who witnessed significant enhancement in the HER performance of NiFeL DH after performing chronopotentiometry (CP) stability test. However,inthe same study,they concluded that before such CP testing,C oFeL DH was the better HER electrocatalyst which in turn was due to the ability of Co in donating electrons to water molecule with al ower applied cathodic potential. Asimilar CoFeLDH fabricated on 3D Ni foam by Babar and co-workers [87] did perform better in terms of overpotential (110 @10mAcm À2 )than the one reported by Rajeshkhanna and co-workers [86] which should have been the results of 3D configuration of the substrate electrode and high mass loading. Realizing the poor HER performance of NiFe LDH in alkali, Zeng and co-workers [88] in their similar work extended to perform total water splitting enriched the NiFe LDH phase with Ni(OH) 2 and observed better activity. Nonetheless,t he performance what they witnessed was still poorer than the ones attained with NiCo and CoFeL DH catalysts.Among the NiFeLDH based HER reports,Guo and co-workers [89] reported as ystematic study by changing the molar ratio of Fe and Ni and the best activity was achieved with the catalyst which had 90 %Feand 10 %Ni, which is very similar to the results that have been achieved with 90 %N i and 10 %C ow ith NiCo LDH reported by Zhao and coworkers. [85] Here,the electron pushing element is Ni just like it . Ab ar diagramofHER overpotentials of Ni(OH) 2 (purple) at 16 mA cm À2 in 1MKOH with various catalytically enhancing (green) and poisoning (red) metal ions and also with metal ions (blue) that had played no significant role in changing its HER activity.All these values were taken from the study of Corrigan and Bendert. [80] was in the case of Co in Ni 0.9 Co 0.1 LDH. However,given that Fe in 3 + state in NiFeLDH is more electronegative than that of Ni in 2 + state in NiCo LDH, the observed difference in HER performance between these two is justified. This fact was actually later supported by the in situ Raman study reported by Qiu and co-workers. [90] In this study,t hey found that with the increasing overpotential (towards the cathodic region), the Raman peak corresponding to FeOOH containing Fe 3+ cation was diminished and led to gradually increasing HER performance as such reduction of Fe 3+ progressed. They also revealed that alkaline HER with NiFeLDH begins with the adsorption of water at Fe 3+ site as it is more electron deficient and proceeded further to form Ni-H intermediate via ac oupled proton-electron transfer (CPET) mechanism. This Ni-H intermediate then could be able to evolve hydrogen either via the Heyrovsky mechanism or Tafel mechanism ( Figure 6). Other than Co and Fe, Cr and La have also been combined with Ni(OH) 2 host to create an active HER electrocatalyst in an alkaline medium of which the NiCr LDH reported by Ye and co-workers [91] showed asubstantially lower overpotential of 138 mV at av ery high current density of 100 mA cm À2 when the ratio of Ni and Cr was 1:0.5. The observed abnormal HER performance was apparently high due to ah uge mass loading (2 mg cm À2 )r ather than the intrinsic catalytic activity changes provided that Cr 3+ does neither donate nor withdraw electrons easily as it is already having ahalf-filled t 2g level (in an octahedral field that exist in LDH phases) which is more stable than electron donating Co 2+ and electron withdrawing Fe 3+ in NiCo and NiFeLDHs, respectively.A nu nusual way of improving the HER activity of Ni(OH) 2 was reported recently by Zhang and Hu [92] which was the doping Ft hat helped them to lower its HER overpotential by 91 mV at benchmarking current density (10 mA cm À2 ). However,the stability test performed for 12 h showed as ignificant increase in overpotential which is attributed to the leaching of Fi na lkaline medium just like any other halide ion does.Hence,this approach did not attract much attention lately to improve the HER activity of other similar HER electrocatalysts.H aving as ynergistic support with aH ER electrocatalyst had always been shown to play benefitting roles.
Conductive and nanostructured carbons (graphene,g raphene oxide,c arbon nanotubes,e tc.) with and without heteroatoms (N,O ,F ,S ,P ,B ,a nd the combination thereof) are such proven synergistic supports in electrocatalysis research. [93] Feng and co-workers [94] have shown for the first time that the HER performance of Co(OH) 2 in alkali can be improved significantly by making ac omposite of Co(OH) 2 electronically conducting polyaniline (PANI) fibres.T his flexible electrode demanded less than 100 mV as overpotential to deliver 10 mA cm À2 for HER. Later, Jia and coworkers [95] achieved better HER performance with NiFe LDH in alkali by compositing it with defect rich graphene sheets.Interestingly,such asynergistic enhancement was also achieved with graphdyine with FeCo LDH by Hui and coworkers [96] which is the only HER electrocatalyst delivered current density as high as 500 mA cm À2 with lower overpotential. However,a st here was no information on the loading of the catalyst, no justifiable comment or correlation can be made with the activity trend being discussed here. Similarly,Bhowmik and co-workers [97] reported HER activity enhancement in the case of CoFeLDH by compositing it with graphitic carbon nitride.T hese reports are among those examples for enhancing the HER performances of hydroxide electrocatalysts with carbon based synergistic supports. Besides compositing with asynergistically enhancing support material, enhancement in HER performances of these binary metal hydroxides was also achieved by defect engineering.
Liu and co-workers [84] were the first ones to introduce this concept of defect engineering in HER electrocatalysis with CoFeL DH just by exfoliation in DMF-ethanol mixture that helped them to realize improved activity.V ery recently,L iu and co-workers [98] also showed that introducing O-vacancyin CoFeL DH could improve its HER activity in alkali. These two studies are thus paving paths to engineer other metal hydroxides discussed above by introducing defects at metal centres or by making O-vacancies which may improve their HER performances as these defects and O-vacancies usually improve the electronic conductivity of these hydroxide materials because of the existence of unsaturated valances at metal and oxygen centres.From the above discussion, it can be concluded here that among all known monometallic hydroxides,C o(OH) 2 is the best HER electrocatalyst and NiCo LDH takes that spot in the category of binary metal hydroxides-based HER electrocatalysts because of the optimal electronic configuration of Co 2+ ion that promotes the Volmer step.
Beyond just mono and binary metal hydroxides,researchers have introduced trimetallic hydroxides as HER electrocatalyst which are sometimes called athird metal doped LDH or directly as LTH. Such ac oncept was first introduced by Wang and co-workers [99] who made aN iCoFeL TH catalyst over carbon fibre cloth substrate that delivered better activity than all other control samples which included NiCo,NiFe, and CoFeLDHs too.Surprisingly,the activity shown for the best binary metal hydroxide (i.e., NiCo LDH) was very low in this report as we are not provided with the loading information here,n or easonable justification can be made.Z hu and coworkers [100] later revealed an optimal composition (Ni 2.5 Co 0.5 Fe)f or the same system that performed relatively better albeit the differences were very small when compared with NiFea nd Ni 2 CoFes ystems.L ater, Babar and co-workers [101] have reported the HER performance of NiCoFe hydroxide system in comparison with that of NiFeL DH (Figure 7a). Though there was as ignificant improvement in overall HER activity,t he retention of the same HER onset overpotential witnessed in this report implied that the observed enhancement was mainly due to the addition of more metal sites rather than intrinsic activity changes.Inthis series of trimetallic hydroxides,Dinh and co-workers [102] have introduced an untouched transition metal of the 3d series,the trivalent vanadium cation (V 3+ ). This is yet another trivalent cation with astable t 2g level (empty) in an octahedral field of aL DH phase just like Cr 3+ which also had as table t 2g level (half-filled) in an octahedral field of aLDH phase.Hence,it did not impose any notable activity change when compared to the activity of NiFeL DH systems discussed above (Figure 7b). This can be witnessed from the same HER onset overpotential that they reported for both NiFea nd NiFeV LDH systems.H ence,t he observed enhancement is mainly because of the added metal sites.T he same trend and observation were made with the introduction of Mo into NiCo LDH systems too which was reported by Hao and coworkers (Figure 7c). [103] All the above-discussed metal hydroxide HER electrocatalysts are benchmarked based on their kinetic current density with respective mass loading wherever provided in Table 1. We urge the readers to be vigilant that this order should not be confused with the trend we discussed here as in some of the reports the loading information was not disclosed and the observed lower overpotential could be an effect av ery high loading of the electrocatalyst.

Metal-Metal Hydroxide Heterostructures
Metals have also been the first choice of electrocatalytic interfaces when it comes to HER in alkali as they were able to provide better electronic conductivity than oxides and hydroxides.M any early works have been concentrated mainly on Pt thin films,n anoparticles,a nd Pt/C composites. [61,62] Other metals such as Ru, Ir, and Rh which showed ac loser overpotentials for HER in alkali were also studied but not carried forward as one of the main objectives of switching water electrolysis from acid to alkali is to avoid these precious metals.
Sheng and co-workers [106] were the first ones to provide ag eneralized activity trend by plotting the exchange current density against the HBE which matched well with the experimental results of many metals (Figure 8a). As anticipated, Pt stood at the apex of the volcano plot with an optimal exchange current density and HBE. Realizing that Pt (111) single crystal performs better than others,S trmcnik and coworkers [107] reported the HER activity trend of Pt (111) (111) surfaces whereas as ignificant loss of activity is witnessed with Mn(OH) 2 and Fe(OH) 2 .T his is justifiable given that Fe 2+ in the low-spin state has completely filled t 2g level and Mn 2+ high spin state has half-filled t 2g and e g levels in an octahedral field making them both stable.I no ther words,b ecause of these stable electronic configurations,Mn 2+ and Fe 2+ ions tend not to donate or withdraw electrons which are necessary for drawing water molecules near hydrogen evolving metal sites. On the other hand, Ni 2+ and Co 2+ have unpaired electrons in their e g orbitals which are essential in deprotonating water molecules by donating them which in turn populate the nearby metal (Pt) site with necessary protons to evolve hydrogen efficiently.B etween Ni 2+ and Co 2+ ,t he former is better for enhancing HER as it does have two unpaired e g electrons.T his also explains the exemplary HER activity of NiCo LDH materials among all other binary metal hydroxides.H owever,t his does not mean that the electronic properties of the metal cations in the metal hydroxide phase alone contribute to alkaline HER enhancement. This was first shown by the work of Subbaraman and co-workers. [56] In this study,b yi ncorporating dissolved Li + cations into Ni(OH) 2 layers that acted as as trong Lewis acid, they were able to achieve af urther weakening of the HÀOH bond in water. Such an additional H À OH bond weakening promoted the water dissociation step by Ni(OH) 2 phase and subsequently enhanced its alkaline HER activity when interphased with Pt (111). Specifically,a65 fold HER activity enhancement was witnessed with Pt(111)/Ni(OH) 2 heterostructure when 0.001 Mo fL i + cations were added to 0.1 MK OH. [56] Hence, both interfacial and electronic effects are concluded to contribute together to enhance the water dissociation step for better alkaline HER. In contrast, the first ones to show such an HER enhancing effect of only Ni(OH) 2 with various other metals such as Ru, Ir,C u, Ag, Au,V ,T i, and Ni and including Pt earlier was Danilovic and co-workers (Figure 8c). [55] In this study,they revealed that irrespective of the

Angewandte Chemie
Reviews metal being studied, the heterostructuring of it with Ni(OH) 2 had always resulted in an improved performance as it facilitates the dissociation of water which is the first and crucial step in alkaline HER. This was further confirmed by the work of Wang and co-workers [108] that showed the possibilities of optimizing the Volmer step by compositing Pt/C with single layer Ni(OH) 2 .A mong all those metals,P t performed better than others and showed parallel activity to that of Ru heterostructured hydroxides and hence,r esearchers were obviously attracted to study the Pt/Ni(OH) 2 system more intensively than others by changing various structural properties of both Pt and Ni(OH) 2 counterparts.Y ua nd coworkers [109] explored the difference in the HER enhancement brought out by the two most common phases of Ni(OH) 2 that exist in alkali in the potential window of HER study.T hey found that the layered alpha phase was poorer than the rocksalt structure having ab eta phase as the latter one improves the kinetics of the Volmer step via afacile water dissociation mechanism. Since it is established that improved electrochemical access to amaximum of Ni(OH) 2 in the Pt/Ni(OH) 2 heterostructure is akey to improve the HER activity further, Yuan and co-workers [110] found that amorphous Ni(OH) 2 decorated with Pt can deliver better HER activity as amorphous materials have always been benefitting water splitting electrocatalysis with improved ECSA. An interesting work of fabricating undercoordinated PtMn alloy and defect rich Ni(OH) 2 heterostructure was made by Wang and coworkers [111] having defects and undercoordination helped them realize significant enhancement in activity.Particularly, the optimal heterostructure was found to have 0.7 wt. %o f defect rich Ni(OH) 2 .Y uand co-workers [112] in arelated work made aP tNi alloy that was heterostructured with Ni(OH) 2 and when the PtNi:Ni(OH) 2 ratio was 1:0.2, the best HER performance was achieved. Beyond just alloying,P tw as structured into different shapes at the nanoscale level as nanoparticles and nanowires and then heterostructured with Ni(OH) 2 by Yina nd co-workers [113] by doing which they witnessed better activity with Pt nanowires/Ni(OH) 2 heterostructure.I tw as Ledezma-Yanez and co-workers [114] who proposed the use of interfacial water reorganization as an appropriate descriptor for Pt/Ni(OH) 2 catalysed HER in alkali using which Sarabia and co-workers [115] later showed the effect of interfacial water structure at Pt(111)/Ni(OH) 2 heterostructure interface by varying the proportion of Ni-(OH) 2 .T hey found that the increasing Ni(OH) 2 proportion increased the HER activity here in this case too.S uch enhancement was even highly pronounced when the flow-cell was used which enabled relatively more access to interfacial water. However,M cCrum and Koper [74] recently showed (as discussed earlier) that more than interfacial water structure, the water dissociation step is crucial in determining alkaline HER activity.Sofar, the combination of Pt and Ni(OH) 2 were only shown to be beneficial in alkaline HER, some of us recently discovered that as imilar enhancement in HER can even be obtained with NiFeLDH by decorating Pt NPs.When the concentration of Pt precursor used to prepare Pt@NiFe LDH heterostructures was increased, the HER activity was also improved (Figure 9a,b). [31] Following this approach, Chen and co-workers [116] reported as imilar HER enhancing with NiFeLDH by decorating Ru NPs instead of Pt. In their  [82] Note:N /A stands for not available and implies that the corresponding data were not provided in the cited reports.

Angewandte Chemie
Reviews case,a ni ncrease in Ru content had led to further enhancement until the wt.% of Ru reached 16 %a nd no significant improvement was witnessed thereafter (Figure 9c,d). In related studies,L ia nd co-workers [117] and Ma and co-workers [118] have recently shown that such HER enhancement can also be witnessed with NiCo LDH and cobalt carbonate hydroxide nanowires using Ru. Next to Pt and Ru, the most studied metal is Cu as ah eterophase with metal hydroxide HER electrocatalysts in alkaline medium. Yu and co-workers [119] showed that encapsulating Cu nanowires with NiFeL DH sheets could improve the HER performance of NiFeL DH significantly but the observed activity is nowhere closer to the ones observed with Pt or Ru interfaced heterostructures.T he same group had also extended this strategy to CoFeL DH and observed as imilar enhancement in the HER activity of CoFeL DH yet not comparable to those of Pt and Ru interfaced ones. [120] Bai and co-workers [121] extended this strategy recently to Co(OH) 2 , Ni(OH) 2 ,a nd as well to CoNi LDH and observed as imilar HER enhancement. Other than Pt, Ru, and Cu, metals that were shown to bring out such HER enhancements were Mn with cobalt carbonate hydroxide,N iw ith Ni(OH) 2 ,P dw ith

Reviews
NiCo LDH, Au with CoNi LDH, and Wm odulated Co-(OH) 2 . [122][123][124][125] Among them, Wm odulated Co(OH) 2 was the only interface that delivered better HER activity than Pt/ C. [123] All these catalysts are benchmarked (in Table 2) by the overpotentials that they demanded at 10 mA cm À2 with different loadings and hence,w es olicit the vigilance of the readers to not to take this order exactly as their actual HER activity trend.  Li et al. [117] Pt j Co(OH) 2 12 93 5.7 Zhou et al. [124] Ru j NiFeL DH N/A 29 31 Chen et al. [116] Ru j Co-(CO 3 )-OH N/A 66 65 Li et al. [118] PtMn j Ni(OH) 2 0.212 75 84.9 Wang et al. [111] Pt j b-Ni(OH) 2 0.0013 92 42 Tang et al. [ Sultana et al. [122] Note:N /A stands for not available and implies that the corresponding data were not provided in the cited reports.
Angewandte Chemie Reviews 7. Metal Oxide/Chalcogenide/Phosphide-Metal Hydroxide Heterostructures In this series of metal hydroxide heterostructures with metal oxides,m ostly LDHs were studied with an oxide heterophase of the same metal ions or with different ones. Theo nly study where am onometallic hydroxide/oxyhydroxide (Co(OH) 2 /CoOOH) was interfaced with an oxide phase of adissimilar metal (PtO 2 of Pt) was reported by Wang and co-workers. [128] In this study,i tw as shown that having Ptbonded Oa tom lowered the water dissociation free energy and improved HER performance.B esides,C oOOH in 3 + state here was actually playing the role of water abstraction site rather than regulating the Volmer step.Later, Wang and co-workers [129] showed that having aC eO x heterophase with NiFeL DH could enhance its HER just by the way in which Co 2+ acted by pushing the electron away from it to lower the energy of water dissociation as Ce in CeO x existed in both 3 + and 4 + states.H owever, the magnitude of HER overpotential lowering is very low when compared to the overpotential lowering brought up by Co 2+ in other systems.A nother example of am etal oxide-metal hydroxide heterostructure with partially different metal centres was reported by Wang and co-workers [130] who employed as eries of hydrothermal and annealing methods to stack NiCo 2 O 4 on Ni foam over which NiFeL DH was grown by using ah ydrothermal reaction. Thee xact intension of this work is still elusive as neither NiFeL DH is ab etter HER electrocatalyst nor Co 3+ ions in the spinel NiCo 2 O 4 is capable of lowering water dissociation energy by donating electrons.H owever,t he observed enhancement in HER activity was significantly higher and we believe that such enhancement could have possibly been brought by the in situ formation of Co 2+ species under the applied reductive overpotentials.N onetheless,n o efforts were made in this work to find out any such in situ presence of HER enhancing Co 2+ species.
However,W ua nd co-workers [131] on the other hand, replaced this spinel NiCo 2 O 4 with an inverse spinel NiFe 2 O 4 and interfaced it with NiFeLDH which had electron donating Fe 2+ ions in the tetrahedral field. As ar esult, as ignificant improvement in HER performance was witnessed. However, the lowering in overpotential was still higher with NiFeLDH-NiCo 2 O 4 catalyst reported by Wang and co-workers. [130] Around the same time,Liand co-workers had shown asimilar HER activity enhancement with NiCo 2 O 4 -NiCo LDH. These studies have shown that not only as econd metal of an appropriate electronic configuration or am etal heterophase but also an oxide heterophase of similar or dissimilar metals that could regulate the Volmer step can bring out notable enhancement in HER activities and imply the importance of

Angewandte Chemie
Reviews having such heterostructures.F igure 10 a-d shows significant HER activity changes noted with such metal oxide-metal hydroxide heterostructures.
Beyond metal oxides,c halcogenides of Co,N i, Fe,M o, and Ww ere also interfaced with monometallic hydroxides and LDHs to realize efficient enhancement HER activity in alkali. Hu and co-workers [132] made first such efficient heterostructure by growing MoS 2 on carbon fiber paper (CFP) followed by hydrothermally sandwiching the same with the best binary hydroxide (NiCo LDH) HER electrocatalysts. Activity enhancement brought up here were mainly due to intrinsic catalytic activity enhancement rather than ac onsequence of ECSA which was indicated by al arge lowering HER onset overpotential and am arginal increase in double layer capacitance (Figure 11 a). Zhu and co-workers [133] in ar elated work modified the approach significantly.I nt his study,t hey exfoliated the established MoS 2 and WS 2 bulk materials and grew various monometallic hydroxides over exfoliated MoS 2 and WS 2 sheets and observed significant intrinsic activity enhancement with all of them among which MoS 2 interfaced with Co(OH) 2 (the best monometallic hydroxide HER electrocatalyst) delivered the best performance (Figure 11 b). Yang and co-workers [134] have later reported as imilar intrinsic activity enhancement with CoNiSe 2 interfaced with CoNi LDH (the best binary hydroxide HER electrocatalyst).
However,w hen either Co 0.85 Se [135] or NiSe [136] were interfaced with ar elatively poor NiFeL DH, the activity enhancement was only marginal and was apparently due to the improved ECSA rather than intrinsic activity enhancement (Figure 11 c,d). In contrast, acontradictory observation with intrinsic activity enhancement with CoSe@NiFeL DH system recently reported by Sun and co-workers [137] stands out of the general trend that we have been showing through this review.Inthese cases of metal chalcogenide-metal hydroxide heterostructures,m etal hydroxide part performs the job of water dissociation (i.e., the Volmer step) and the metal chalcogenide surface catalyse the HER by forming hydridic intermediates.I nt his category of metal chalcogenide interfaced metal hydroxide,w eh ave very recently reported an Figure 11. (a,b) Intrinsic HER activity enhancement realized by heterostructuring the best LDH (NiCo) and monometallic hydroxide (Co(OH) 2 ) HER with metal chalcogenides reproduced,r espectively from ref. [132] (Copyright 2017, Cell Press) and ref. [133] (Copyright 2018, Wiley). (c,d) HER enhancement witnessed because by heterostructuring CoSe and NiSe with NiFeLDH mainly by improvedE CSA rather than intrinsic activity enhancement (indicated by the same HER onset overpotentiala nd EG stands for exfoliated graphite) reproduced,r espectively from ref. [135] (Copyright2 016, RSC Publishing) and ref. [136] (Copyright2 017, ACS). efficient way of making such heterostructures with an unconventional way. [49] We subjected the NiS sheets grown hydrothermally over Ni foam substrate for anodic hydroxylation to form Ni(OH) 2 heterophase rapidly.T his method has additionally benefitted the HER activity improvement by increasing ECSA as ar esult of surface amorphization. As aresult of collective benefit from both heterostructuring and amorphization, more than 100 mV of overpotential lowering was witnessed in alkaline HER at 100 mA cm À2 .T his method provides af acile path for heterostructuring other non-oxide/ hydroxide materials that can perform better in alkaline HER.
Metal phosphides interfaced metal hydroxide heterostructures are also reported to be aclass of highly active HER electrocatalysts next to metal oxides and metal chalcogenides interfaced metal hydroxide heterostructures.Z hou and coworkers [138] had shown that making of CoNiP-NiFeL DH could result in as ubstantial HER activity enhancement in alkali albeit the magnitude of enhancement was relatively low as expected because NiFeL DH is neither ag ood binary hydroxide HER electrocatalyst nor acocatalyst that facilitate the Volmer step of the reaction. Xiong and co-workers [139] have purposefully formed al ayer of Ni(OH) 2 over Ni 2 P thereby forming ah eterostructure witnessed as ignificant HER activity enhancement. They also revealed that the time of hydrothermal treatment and the associated thickness of Ni(OH) 2 layer could influence the activity to agreater extent. Their results showed that a10hhydrothermal treatment was optimal in attaining the best HER enhancement out of this system. Zhang and co-workers [78] first showed that Co 2 Pw as prone to surface hydroxylation even under reductive potentials that degraded the activity.H owever,S ua nd co-workers [140] reported an important finding that anodizing CoP with 20 mA cm À2 could form undercoordinated Co(OH) x species over CoP that helped them to realize ab etter HER activity.
Like the metal chalcogenide interfaced metal hydroxide systems,m etal hydroxides interfaced with metal phosphides (which are regarded as metallic in nature due to their high electronic conductivity) do also follow the same mechanism of HER where the hydroxide phase facilitates the water dissociation step and metallic metal phosphide perform HER by forming hydridic intermediates.A ll these catalysts are benchmarked taking the overpotentials that they demanded at 10 mA cm À2 in Table 3a nd we also do remind here again that the readers should not adopt this order of benchmarking unless otherwise they are convinced with the loading of the catalyst material, ECSA, size,shape,and morphology.
All the above benchmarking tables (Table 1-3) display the catalysts in ascending order of the overpotentials that these catalysts demanded at 10 mA cm À2 .T his geometrical area normalized current density is highly dependent on the loaded mass of the catalyst. [64] Loading is an important piece of information in all sorts of electrocatalytic studies including this HER which unfortunately is not disclosed in most of the reports discussed in this report.
This implies that the overpotential based benchmarking could be potentially wrong and misleading.Hence,tohave an alternative view on the activity trends among all these catalysts,w ep ortrayed all these catalysts discussed in each category according to their Tafel slopes ( Figure 12) which primarily indicates the mechanism of HER and rate determining step (RDS). Since water dissociation coupled proton discharge is the RDS (i.e., Volmer step) of HER in alkaline medium and the Tafel slope of this step should in the range of 50-120 mV dec À1 ,a ny electrocatalyst that is having aT afel slope 50 mV dec À1 can be considered ab etter active electrocatalytic interface.W eh ave also shown in one of our recent studies that loading tends to affect Tafel slopes too but not as extensively as it does to the geometrical area Table 3: Benchmarking of metal oxide-metal hydroxide, metal chalcogenide-metal hydroxide, and metal phosphide-metal hydroxide heterostructures based HER electrocatalysts as per the overpotentials they demandedf or driving 10 mA cm -2 in 1.0 MK OH.

Angewandte Chemie
Reviews normalized current densities and the corresponding overpotentials determined from it. [64] From Figure 12, it is clearer that the Volmer step of alkaline HER is promoted highly with metal-metal hydroxide heterostructures than other discussed ones and the pristine hydroxides.S pecifically,t he heterostructures of Pt, Ru, and Cu were shown to have the lowest possible Tafel slopes for HER in alkali. Therefore,w e conclude here with the knowledge we acquired via an extensive literature survey that both metal NPs decorated metal hydroxide or metal hydroxide clusters deposited metal substrates can be efficient alkaline HER electrocatalysts. These are the ones that are so far shown to have demonstrated lower overpotentials and Tafel slopes in addition to better stability and selectivity.

Understanding Alkaline HER on Different Interfaces
It is now certain that the efficiency of alkaline HER is greatly dependent on the efficiencyofwater dissociation into proton and hydroxide at the cathode which concurrently forms the hydridic intermediate with the catalytic site.T his is the typical Volmer step of alkaline HER. This step of the reaction is so difficult as it produces more hydroxide at the interface which is already populated densely with it. Hence, the existence of ahigh concentration of hydroxide ions on the cathode surface pushes the water dissociation equilibrium backwards.F rom our ongoing discussions on the HER activity trends observed with metal hydroxides and their heterostructures,ag raphical sketch (Scheme 2) is drawn illustrating which part of the catalyst tends to initiate water dissociation and which is efficient in making hydridic intermediate which then can evolve hydrogen molecule by following either Heyrovsky or Tafel mechanism. [54] On the surfaces of metals that do not form any oxide or hydroxide under applied cathodic potential (e.g.P ta nd Ru), water dissociation step is highly energy demanding and hence require high overpotentials to perform HER in order to achieve the desired current density.Asaconsequence of the fact that both water dissociation and hydridic intermediate formation ought to occur on the same surface metals usually perform poorly in alkaline medium. This is also the reason behind observing Tafel slope as large as 120 mV dec À1 and even higher. Metal hydroxide surfaces,onthe other hand, are capable of facilitating the tedious water dissociation step because of their ability to coordinate reversibly with water and hydroxide ligands as both of them occupy closer places in the spectrochemical series.A saresult, water molecules that coordinate with am etal cation site in am etal hydroxide surface can easily break the OÀHb ond of water when compared to the noble metal surfaces.This facileness of water dissociation with metal hydroxide surface enriches it with short living protons which can immediately undergo discharge to form the hydridic bond with anearby catalytic site.
This catalytic site can be another metal cation in the same metal hydroxide phase or metals (Pt, Ru, Cu, Pd, W, Pd, etc.,), metal chalcogenides (of Ni, Co,M o, W, etc.,), and metal phosphides (of Ni, Co,M o, etc.,) that are interfaced with them. Among pristine metal hydroxides,Co(OH) 2 is shown to be the better one as it dissociates water and concurrently forms hydridic bond by donating e g electron(s). In binary metal hydroxides,N iCo LDH is the best as they both have unpaired e g electrons that facilitate both water dissociation and hydridic bond formation by "pushing" mechanism according to Zhao and co-workers. [85] When it comes to heterostructures of metal hydroxides,N i(OH) 2 heterostructured with Pt and Ru is the best as Ni(OH) 2 splits O À Hbond in water and the metallic counterpart (Pt or Ru) forms the hydridic intermediate concurrently which in turn facilitate HER. LDHs and other multi-metallic hydroxides are also shown to exhibit similar activity trends when they are interfaced with Pt and Ru. However,C u-interfaced metal hydroxide and LDH heterostructures get benefitted from the better electronic conductivity of Cu. Hence,t hey impart better enhancement when they are sandwiched or wrapped around with as uitable metal hydroxide or LDH co-catalyst. This is because metallic Cu tends to form Cu + and Cu 2+ species at potentials that are closer to HER and HOR when exposed directly to the electrolytes which cannot catalyse the Heyrovský and Tafel reactions efficiently.B esides,o xygen containing Cu + and Cu 2+ are well-known for their poor HER activity which is one of several reasons why they are attractive for CO 2 reduction. [142] In av ery recent systematic study by Boettcher and co-workers, [143] the predominance of water dissociation with NiO (which can form Ni(OH) 2 in alkali)i s ascertained once again to be controlling alkaline HER efficiency.T his was revealed from the relationship between water dissociation overpotentials and the point of zero charges (pzc) of various water dissociating catalysts placed on the locally alkaline side of ab ipolar membrane (BPM) (Figure 13). Catalysts with acidic and neutral pzc values did not show any rationalizable relationship with the water dissociation overpotentials.H owever,i na lkaline conditions, NiO oxide with the highest pzc had the lowest water dissociation overpotential when compared to Fe(OH) 3 , Co 2 O 3 ,a nd Al 2 O 3 .I nt he case of metal chalcogenides and phosphides interfaced metal hydroxide heterostructures,t he Scheme 2. Illustration of water dissociation step's (Volmer step) site preferences on different interfaces in alkaline medium.

Angewandte Chemie
Reviews hydroxide counterpart is predicted to perform the water dissociation step because of its ability to exchange water and hydroxide ligands easily.
It is recently comprehended that all non-oxide/hydroxide HER electrocatalysts that were previously reported to be active in the alkaline media were actually forming al ayer of hydroxide on their surface which in turn contributed significantly to the overall catalytic process. [144][145][146] Perceiving this, the research community has now started engineering the surfaces of such non-oxide/hydroxide alkaline HER electrocatalysts purposefully to impart better activity.Examples are our recent work on NiS/Ni foam [49] and the one by Su and coworkers [140] in which we employed anodic sweeping technique whereas they employed galvanostatic anodization that resulted in undercoordinated Co(OH) x on CoP surfaces. Hence,i ti sc oncluded here that having am etal hydroxide heterophase with metal ions of appropriate electronic configurations (preferably with unpaired e g electrons) with known metal/metal chalcogenide/metal phosphide catalyst is the key to achieve better efficiency in alkaline HER.

Future Directions of Alkaline HER
As it is understood now that water dissociation plays akey role in determining the efficiencyo fa na lkaline HER electrocatalyst, several works have now been focused on this particular part of the overall HER electrocatalysis. Meanwhile,i ti sa lso important to concentrate on the Heyrovsky and or Tafel steps.A lthough these two steps of HER primarily depend on the Volmer step,the bond strength of hydridic intermediate formed with the catalytic site also matters significantly.Inour earlier review, [8] we have provided an elaborated overview on the bond strengths of Htovarious elements that compose most of the known HER electrocatalysts. [8] Forafacile Heyrovsky and or Tafel step,i ti s essential to have amoderate SÀHbond strength. In this point of view,P ta nd Ru from the metals and Pa nd Se from nonmetallic counterparts of the metal chalcogenide and phosphide catalysts have moderate strength. Hence,w ee xpect that future research will be concentrated on these elements for quite al ong time until af urther breakthrough occurs.A list S À Hb ond strengths can be found elsewhere for the appropriate design of HER electrocatalysts. [147][148][149] In situ and operando spectroelectrochemical characterizations have been of tremendous use recently in understanding the structure of catalysts under operating conditions at various applied potentials and also in tracking the structure of electrical double layer and the species that exist during electrocatalysis. [44] Since alkaline HER electrocatalysis is widely accepted to begin with the water dissociation step that provides the necessitated protons at the cathode surface, it is also essential to study the interface and electrical double layer simultaneously to know more about this widelyaccepted mechanism. An interesting and insightful study was reported recently by Qiao and co-workers [150] on the HER electrocatalysis of Pt in alkali using in situ Raman spectroscopy.T hey found that when Pt/C was subjected to catalyse HER at high overpotentials,d ue to av igorous water dissociation, ah ighly acidic localized environment in the vicinity of the electrode was generated which successively increased HER activity.T he same was confirmed using deuterated water and alkaline medium too (Figure 14 a-c). This effect was observed only with the nanostructured Pt/C but not with the bulk Pt indicating that the size and shape of electrocatalysts can have huge impacts on the way in which HER is performed. This study also implies that such in situ or operando spectroelectrochemical studies of alkaline HER with all the above-discussed metal hydroxide catalysts and their heterostructures may even lead us to achieve fascinating results and could improve our understanding.M eanwhile, they could also prove our current understanding of alkaline HER be incorrect and provide other exciting insights.Hence, we sincerely hope that these analytical tools will play critical roles in the advancement of alkaline HER electrocatalysis. Other than this,t he performance of every catalyst under alarge A/V condition (large electrode surface area (A) with smaller electrolyte volume (V)) at elevated temperature which is the typical industrial condition may vary significantly. [151] Hence,i tw ill be inevitable to evaluate all these catalysts under large A/V conditions too.
As the availability of water molecules at the interface is crucial, flowing electrolyte will refresh the surface with anew electrolyte solution and will lower the chances of additional basification as aconsequence of water dissociation which has the tendency to lower the HER kinetics.H ence,t he mesoscopic effects of water electrolysers must be taken into account in the evaluation stage too. [20] With research works that are yet to be done and reported in the above-mentioned directions,i ti sc ertain that alkaline water electrolysis Figure 13. The relationship between water dissociation overpotentials at 20 mA cm À2 and point of zero charges (pzc) of various water dissociating catalysts placed on the alkaline side of abipolar membrane showing alinear inverse relationship and signifying the importance of water dissociation step in efficient alkaline HER. Reproduced from ref. [143] (Copyright 2020, AAAS).
(especially the HER part) can be made energy-efficient in near future.

Summary and Outlook
Unaffordability of acidic OER electrocatalysts led to the development of alkaline water electrolysers with earthabundant OER electrocatalysts albeit that came at ap rice of compromised HER activity in alkali. Absence of free protons in alkali made the relatively simple and straightforward HER tedious and complex even with the state-of-the-art electrocatalysts.I na lkaline HER, to get the necessary localized proton rich environment, it is essential to cause the water dissociation step prior to the classic Volmer step of HER. Hence,i na na lkaline medium, the water dissociation and concurrent proton discharge that form hydrogen evolving hydridic intermediates are considered as the Volmer step which controls the overall efficiency of agiven electrocatalyst to ag reater extent. From recent attempts of using metal hydroxides and their heterostructures as HER electrocatalysts in alkaline media, it is now understood that the metal hydroxide phase facilitates the water dissociation. This is because of their ability to reversibly coordinate with both water and hydroxide ligands which are positioned closer to one another in the spectrochemical series.T his has been the reason for all the enhanced HER activity witnessed with Pt-Ni(OH) 2 and other related systems.Whenthere is only metal hydroxide,t he same metal centres which perform water dissociation do also form hydrogen evolving hydridic intermediates.H owever,t hey are relatively stronger in bond strength than the ones formed with am etal atom or the chalcogenide/phosphide anion site in their heterostructures. As aconsequence of this,metal hydroxides alone were poor in Figure 14. In situ Raman spectra acquired with Pt/C in water-alkalinem edium (a), with bulk Pt in water-alkalinem edium (b), and with Pt/C in the deuterated water-alkalinemedium. Existenceo fhydronium ion/protonated deuterium water or the combination thereof were witnessed only with Pt/C. reproduced with permissionf rom ref. [150] (Copyright 2020, NPG).

Angewandte Chemie
Reviews catalysing HER in alkali. However,i ft he electronic configuration of the metal ions in the metal hydroxide catalyst being studied is appropriate (i.e.h aving unpaired e g electrons to readily donate and dissociate water) and the hydridic intermediate formed with them are of optimal bond strength, they could perform relatively better than other metal hydroxides.A ne xample is Co(OH) 2 with ac onductive carbon support. Having understood that interfacial water structure and its dissociation control the overall efficiency of alkaline HER, we are in an urge to study not only the electrode surface but also the interfacial water structure, electrical double layer, and their reorganization while varying the applied potentials.T his can give am ore valuable understanding of alkaline HER and the way it does work. This is expected to be accomplished via several in situ and or operando spectroelectrochemical studies to be done in the future.Besides,concentrating on the mesoscopic effect of the electrode and cell is also equally important under large A/V condition to truly realize ab etter HER performance under industrial operating conditions.W ith this,w ec onclude here that alkaline HER will step into anew era of its development in the upcoming years provided that the predicted directions of research in this field will lead so.Eventually,this may end the quest for an energy-efficient alkaline HER with affordable electrocatalysts (metal hydroxides and their heterostructures) that outperformed Pt/C already at all overpotentials.