Computational and experimental studies of a Ni/Pt bimetallic catalyst for H2 production from ammonia decomposition

2009 Spring Symposium

 
Danielle A. Hans­gen
Depart­ment of Chem­i­cal Engi­neer­ing
Uni­ver­si­ty of Delaware
Newark, DE


Abstract — The ammo­nia decom­po­si­tion reac­tion has recent­ly received increased atten­tion due to the pos­si­bil­i­ty of ammo­nia being used as a hydro­gen stor­age medi­um in a pos­si­ble hydro­gen econ­o­my. We have explored this decom­po­si­tion reac­tion through mul­ti­scale micro­ki­net­ic mod­el­ing for a num­ber of tran­si­tion met­al cat­a­lysts, includ­ing Cu, Pt, Ir, Ru, Pd, Rh, Co, Ni, Fe, W, and Mo, to bet­ter under­stand the reac­tion mech­a­nism. An under­stand­ing of the reac­tion mech­a­nism and elec­tron­ic prop­er­ties of these met­als has giv­en insight into how to tai­lor cat­a­lysts to improve cat­alyt­ic activ­i­ty for this reac­tion.

The mech­a­nism con­sists of 12 ele­men­tary reac­tion steps and 5 sur­face species, name­ly N, H, NH, NH2, and NH3. For many of the met­als, a large por­tion of the sur­face is cov­ered by adsor­bates. For these met­als, repul­sive adsor­bate-adsor­bate inter­ac­tions were expect­ed to change the bind­ing ener­gies of the sur­face species, there­by chang­ing the ele­men­tary reac­tion acti­va­tion bar­ri­ers and mod­i­fy­ing the cat­alyt­ic activ­i­ty [1]. Cov­er­age depen­dant atom­ic heats of chemisorp­tion were cal­cu­lat­ed through DFT using the Vien­na Ab-ini­tio Sim­u­la­tion Pack­age (VASP) for the var­i­ous tran­si­tion met­al cat­a­lysts. Cov­er­age depen­dant mol­e­c­u­lar bind­ing ener­gies were cal­cu­lat­ed using a method based on scal­ing rela­tion­ships pub­lished by Abild-Ped­er­son et al. [2] and acti­va­tion bar­ri­ers were cal­cu­lat­ed through the bond-order con­ser­va­tion (BOC) method [3].

Inclu­sion of the inter­ac­tion para­me­ters to the mod­els result­ed in reduced nitro­gen cov­er­ages and a peak shift in the vol­cano curve. The con­ver­sions were plot­ted against the char­ac­ter­is­tic nitro­gen heat of chemisorp­tion for each met­al, which was found to be an ade­quate descrip­tor for this reac­tion. The vol­cano curve of the con­ver­sions cal­cu­lat­ed through the micro­ki­net­ic mod­els are in good agree­ment with exper­i­men­tal data of sin­gle met­al cat­a­lysts by Gan­ley and cowork­ers [4]. The max­i­mum activ­i­ty was found at a nitro­gen heat of chemisorp­tion of approx­i­mate­ly 130 kcal/mol.

A DFT study of nitro­gen bind­ing ener­gies on Pt-3d bimetal­lic sur­faces showed a bind­ing ener­gy of 131 kcal/mol on the Ni-Pt-Pt sur­face, indi­cat­ing that it could be a poten­tial­ly active cat­a­lyst; there­fore sur­face sci­ence exper­i­ments were per­formed to assess the micro­ki­net­ic mod­el and DFT results. The Ni-Pt-Pt sur­face was found to be more active at decom­pos­ing ammo­nia at low tem­per­a­tures and des­orbed nitro­gen at low­er tem­per­a­tures than a Ru(0001) sur­face [5], cur­rent­ly the most active sin­gle met­al cat­a­lyst

Speaker’s Biog­ra­phy — Danielle Hans­gen received her Bachelor’s degree in chem­i­cal engi­neer­ing in 2005 from the Uni­ver­si­ty of Wash­ing­ton. She is cur­rent­ly a third year, PhD can­di­date in chem­i­cal engi­neer­ing at the Uni­ver­si­ty of Delaware. She is advised by Dr. Dion G. Vla­chos and Dr. Jing­guang G. Chen and is work­ing on the ratio­nal design of cat­a­lysts for the ammo­nia decom­po­si­tion reac­tion.