= Solution
Use the paper's classical norm convention: $|u|_{2;U}$ is the <derivative supremum norm> through order two, not an $L^2$ norm. Write $H(u;U)=[D^2u]_{\alpha;U}$ and $Q(u,f)=|u|_{2;B_1}+|f|_{0,\alpha;B_1}$, where the latter is the full <Hölder norm> of the forcing.
Fix $0<\delta<1$. If the asserted estimate failed, rescaling the functions by their global <Hessian matrix> <Hölder seminorm> would produce $u_j,f_j$ with
$$
H(u_j;B_1)=1,\qquad H(u_j;B_{1/2})>\delta+jQ(u_j,f_j).
$$
In particular $Q(u_j,f_j)\to0$. Choose $x_j,y_j\in B_{1/2}$ for which
$$
|D^2u_j(y_j)-D^2u_j(x_j)|>\tfrac\delta2|y_j-x_j|^\alpha.
$$
The <supremum norm> of $D^2u_j$ tends to zero, so $r_j=|y_j-x_j|\to0$. Let $P_j$ be the quadratic <Taylor polynomial> of $u_j$ at $x_j$, and define
$$
U_j(z)=\frac{u_j(x_j+r_jz)-P_j(x_j+r_jz)}{r_j^{2+\alpha}}.
$$
The domains contain $B_{1/(2r_j)}$, and $U_j,DU_j,D^2U_j$ vanish at zero. Moreover
$$
[D^2U_j]_{\alpha}\leq1,\qquad
\Delta U_j(z)=r_j^{-\alpha}\bigl(f_j(x_j+r_jz)-f_j(x_j)\bigr).
$$
The right side is bounded by $[f_j]_\alpha|z|^\alpha$ and tends to zero locally uniformly. The normalized <Hessian matrix> bound controls $|D^2U_j(z)|\leq|z|^\alpha$; integration along line segments then bounds $DU_j,U_j$ on every compact <Euclidean ball>. <Arzela-Ascoli theorem> and a <diagonal subsequence argument> give convergence in $C^2$ on compact subsets to an entire <harmonic function> $U$, with $[D^2U]_{\alpha;\mathbb R^n}\leq1$ and $D^2U(0)=0$.
After a further subsequence $z_j=(y_j-x_j)/r_j\to z_*$ with $|z_*|=1$. The normalization ensures $|D^2U(z_*)|\geq\delta/2$. But each second derivative of $U$ is harmonic and has finite global <Hölder seminorm> with exponent below one. The supplied Liouville theorem makes each derivative constant; it is the globally Hölder case of the <polynomial-growth Liouville theorem for harmonic functions>. Its value at zero makes it zero, a contradiction. This <blow-up compactness proof of an interior Schauder estimate> establishes
$$
\boxed{H(u;B_{1/2})\leq\delta H(u;B_1)+C_{n,\alpha,\delta}Q(u,f).}
$$
To absorb the term on the larger <Euclidean ball>, rescale this estimate to <Euclidean balls> contained in $B_s$. For $1/2\leq r<s<1$, pairs separated by less than $(s-r)/2$ are controlled in a <Euclidean ball> centred at their first point; farther pairs are controlled directly by $\|D^2u\|_\infty$. Thus
$$
H(u;B_r)\leq\delta H(u;B_s)+C_\delta(s-r)^{-2-\alpha}Q(u,f).
$$
This is the covering step in the <Simon absorption lemma>. Take $r_j=1-2^{-j-1}$ and fix $\delta<2^{-2-\alpha}$. Iterating makes the remainder $\delta^jH(u;B_{r_j})$ tend to zero, since the global seminorm is finite; the error terms form a convergent geometric series. Consequently
$$
\boxed{[D^2u]_{\alpha;B_{1/2}}\leq C_{n,\alpha}\bigl(|u|_{2;B_1}+|f|_{0,\alpha;B_1}\bigr).}
$$
For the requested failure of the weaker estimate, take $u_m=m^{-2}\sin(mx_1)$ and $f_m=-\sin(mx_1)$. The <derivative supremum norm> $|u_m|_2$ and $\|f_m\|_\infty$ remain bounded, but comparison at $0$ and $\pi e_1/(2m)$, both in $B_{1/2}$ for large $m$, gives
$$
\boxed{[D^2u_m]_{\alpha;B_{1/2}}\geq(2m/\pi)^\alpha\longrightarrow\infty.}
$$
This proves the <necessity of Hölder forcing for Schauder estimates>. Every member of the sequence is smooth; failure comes from the increasing frequency, not from any lack of individual regularity.
Back to article page
