We thank the referees for their positive and helpful comments. We also
appreciate their detailed corrections of the text. We have modified
the manuscript according to their suggestions. The changes are
outlined below:

Comment:
The paper is slightly longer than the page limit of 6 pages set by the conference organisers.
I think the use of the phrase "packing of the molecules" is sometimes not ideal. The authors essentially show that the field distributions are not dependent on the type of lattice or the amount of (random) displacements of the SMMs from their lattice positions (see Figures 2-4) and argue that the stacking of the molecular layers "will not change [their] general conclusions". To me it would therefore seem that the "packing" of the molecules is not significant, but only the "SMM configuration", i.e. the orientation of the moments, as nicely outlined in the authors' discussion section. I propose to amend "SMM packing" to "SMM configuration" or "SMM moment configuration" in the appropriate places.

Reply:
We have changed "packing" to "configuration" where appropriate.

Comment:
Abstract:
missing words in "... field distribution (acting?) on a single..." and "... by tuning (the) molecule-substrate..."
"...packing of the molecules...": I suggest "configuration of the molecule moments" for the reasons outlined above.

Reply:
We have followed the referee suggestion and made the changes accordingly.

Comment:
Introduction:
"As a result(s), a considerable..."
"...for depth-resolved [22] investigation(s)..."
"...SMM packing/configuration...": "... SMM configuration..."
"...lower fluctuation rate(s) is expected..."
"...enhanced dynamics (are) expected far away..."
"We point out here (,) that SMMs..."
"...other factors as well(,) which we do not consider here."

Reply:
We have corrected the text accordingly.

Simulations:
"Near the substrate(,) the molecular..."
a few issues with vector notation:
i)mu is not given the bold vector symbol, despite being equated to the bold vector symbol for spin S
ii) equation (1) contains different fonts for the subscripts "i" and the symbol mu needs bolding to make it a vector
iii) I think equation (1) needs two separate indices, i.e. the field acting on moment mu_i due to moment mu_j
"where r_i is the (spacing-> vector) between them."
"mu=1mu_B": should be "|\boldsymbol{\mu}|=\mu_B" I think? (so you assume g=2 and S=1/2, right?)
Please include information on what values you used for the cutoff distance rho (in terms of lattice paramater "a" I guess) and the lattice parameter "a", since you later give absolute distance values in Angstrom rather than "a" without supplying the value of "a".

Reply:
We have corrected the text accordingly. We do not think that the
equation needs two indices since the field acting on a certain SMM
does not depend on its moment, only on its neighbours (one index).

The cutoff value is now added to the text (\rho <= 5a). We also added
a note that we use a cutoff value to acheive convergence of the
calculated distribution.

Comment:
Results:
3.1. Near the substrate:
"...will broaden the distribution..."
Please comment on two things concerning p(B_parallel=0):
i) Why is it lower in the plot (~5.5%) in Figure 2 than the 31.25% you give in the text (I imagine it is due to next nearest neighbours?) for the triangular lattice.
ii) If I repeat the calculation for the square lattice I get a higher p(B=0)=6/16=37.5%, but in Figure 2 the square lattice has a lower p(B=0) than the triangular lattice. Why?
"However, different stacking will not change our general conclusions." Could you elaborate on this a bit? I guess this ties in with the later investigation of the randomized lattice?

Comment:
3.2. Far fro the substrate:
"It also (This?) remains unchanged when..."

Reply: 
Corrected.

Comment:
3.3. Random lattice:
Are you displacing all the SMMs or just the one that you calculate the B-fields at? The phrase "random displacement of the SMM from the lattice sites" is ambiguous since SMM is singular while lattice sites is plural.
Please mention for which lattice type (looks like triangular) you did these calculations and how many layers you included.
"...and phi a random angle(s)..."
"The features, due to the nearest neighbours, mentioned above..."

Reply:
The typos were corrected.
The displacement is for all SMMs (plural) and is now more clearly
stated in the text.
The calculation is done only for the triangular lattice as an
example. However, the results are qualitatively the same for the
square lattice. Note tnat with increased randomness there is no
difference between triangular and square lattices. This is now more
clearly stated in the text.

Comment:
Discussion:
"...causes a (mixing) of eigenstates..."
"A system(s) (,) which is initially..."
"...fields along the easy axis reduce tunnelling (and) slow(s) spin dynamics ..., while field(s) perpendicular..."
"...we compare spin (dynamics) in the two ..."
"...are identical with a Gaussian (of width) ..."

Reply:
The typos are correctd.

Comment:
What value for the lattice parameter "a" and what lattice type and how many layers did you use to calculate the Gaussian widths?
Is there a possible explanation why "these general conclusions do not change with the type of lattice or the amount of random disorder" (i.e. the "packing" of the molecules) in both scenarios?

Reply:
The value of "a" is not relevant for the calculations as described in
the "Simulations" section.

Regarding the "general conclusions" we meant that the effects of
B_parallel and B_perp on spin dynamics does not change significantly
by changing the lattice type or introducing random disorder. This is
since the field distribution is primarily affected by the orientation
of the neighboring spins rather than their geometric configuration. We
added these clarifications to the text.

Comment:
Conclusion:
"...experienced by an individual SMM (,) along ..."
"molecular packing/configuration": again I think just "molecular configuration" would be better
"...spin dynamics to change(s) as a function..."
"...layers near the (substrate) interface,..."

Reply:
The typos are corrected.

Comment:
References:
Some of the references contain too many author names. From the IOP template file: "Up to ten authors may be given in a particular reference; where there are more than ten
only the first should be given followed by ‘et al’."
In the main text of the paper the reference numbers in square brackets are sometimes given as a long list of individual numbers. I think something more like [7-18] would be better for example in the Introduction section.

Reply:
We have corrected according to IOP guidelines. The separate references
is according to IOP guidelines.