Improve performance of computing molecular integrals and Hamiltonian

[soranjh]

Context:
The PR improves the performance of computing the electron repulsion integral and simplifying the qubit Hamiltonian.

Description of the Change:

• The electron_repulsion function contained repeated calls inside nested for loops. These steps are now taken out of the oops and are computed only once.
• The _hermite_coulomb function also contained a loop for computing the Boys function which is now removed and the Boys function is called once for evaluating an array. The _boys function is modified to adopt with this change.
• The simplify function has more efficient ways for accessing the Hamiltonian terms..

Benefits:
The efficiency of building a molecular Hamiltonian is improved significantly.

Possible Drawbacks:
NA

Related GitHub Issues:
NA

[github-actions]

Hello. You may have forgotten to update the changelog!
Please edit doc/releases/changelog-dev.md with:

• A one-to-two sentence description of the change. You may include a small working example for new features.
• A link back to this PR.
• Your name (or GitHub username) in the contributors section.

[codecov]

Codecov Report

Merging #2316 (b2ca3fe) into master (75e4c40) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master    #2316   +/-   ##
=======================================
  Coverage   99.42%   99.42%           
=======================================
  Files         243      243           
  Lines       18322    18324    +2     
=======================================
+ Hits        18216    18218    +2     
  Misses        106      106           

Impacted Files	Coverage Δ
pennylane/hf/hamiltonian.py	100.00% <100.00%> (ø)
pennylane/hf/integrals.py	100.00% <100.00%> (ø)
qchem/pennylane_qchem/qchem/hf/integrals.py	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 75e4c40...b2ca3fe. Read the comment docs.

[Jaybsoni]

Pretty straight forward, just a few clarifications.

Thanks,

[obliviateandsurrender]

Looks good @soranjh!

[Jaybsoni]

Looks good to go

[ixfoduap]

Just double-check if using t == 0 or t == 0.0 is better. Otherwise good to go! Beautiful PR identifying a simple and very impactful change